Aleuts, Sea Otters, and Alternate Stable-State Communities
Author(s): Charles A. Simenstad, James A. Estes, Karl W. Kenyon
Source: Science, New Series, Vol. 200, No. 4340 (Apr. 28, 1978), pp. 403-411
Published by: American Association for the Advancement of Science
Stable URL: http://www.jstor.org/stable/1746443
Accessed: 21/12/2009 02:13

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=aaas.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.




        American Association for the Advancement of Science is collaborating with JSTOR to digitize, preserve and
        extend access to Science.




http://www.jstor.org
                                               such as sea urchins (Strongylocentrotus
                                               polyacanthus) (14), limpets (Collisella
                                               pelta), and chitons (Katharina tunicata,
                                               Cryptochiton stelleri) to sparse popu-
                                               lations of small individuals. This inter-
                                               action in turn allows an abundant associ-
       Aleuts, Sea Otters, and                            ation of macroalgae to flourish on the
                                               rocky substrate of the broad littoral
  AlternateStable-StateCommunities                             benches and shallow (0 to 20 meters)
                                               sublittoral zones (7, 10). In contrast, is-
                                               lands with few or no sea otters support
      Charles A. Simenstad, James A. Estes, Karl W. Kenyon               dense populations of large herbivorous
                                               invertebrates which, by overgrazing, vir-
                                               tually exclude the association of fleshy
                                               macroalgae. These islands are character-
  Interpretations of paleoecological evi-  community structures are maintained by     ized by bare rocky substrates covered by
dence in the Aleutian Islands have been    the presence or absence of sea otters in    a dense carpet of sea urchins and, in
made with the assumption that aboriginal   the Aleutian Islands (7), supporting      some areas, abundant bivalves (Modi-
Aleuts exploited and maintained a stable   Sutherland's (8) evidence that multiple    olus rectus), colonial tube worms (Pota-
and uniform resource base (1, 2). Laugh-   stable-state communities can occur in     milla reniformis), predaceous asteroids
lin (2) supposed the ecological role of    one environment. Our intent here is to     (Leptasterias alaskensis,   Crossaster
aboriginal Aleuts to be "a moderating in-   integrate this understanding of sea otter-  papposus, Solaster stimpsoni and a num-
                                              ber of species yet to be identified), epi-
                                              benthic macrocrustaceans (Telmessus
  Summary. Reexamination of stratified faunal components of a prehistoric Aleut      cheiragonus, Erimacrus isenbecki, and
midden excavated on Amchitka Island, Alaska, indicates that Aleut prey items        Elassochirus tenuimanus), and octopus
changed dramatically during 2500 years of aboriginal occupation. Recent ecological     (Octopus dofleini) (15).
studies in the Aleutian Islands have shown the concurrent existence of two alternate      The association of macroalgae is the
stable nearshore communities, one dominated by macroalgae, the other by epiben-       major source of marine primary produc-
thic herbivores, which are respectively maintained by the presence or absence of      tion in the western Aleutian Islands and
dense sea otter populations. Thus, rather than cultural shifts in food preference, the   other north temperate areas (16). Con-
changes in Aleut prey were probably the result of local overexploitation of sea otters   sequently, islands lacking sea otters (and
by aboriginal Aleuts.                                    thus the robust association of macroal-
                                              gae) apparently are relatively unproduc-
                                              tive compared with islands where sea ot-
fluence on population fluctuations in the  induced alternate communities with a     ters are abundant (7, 17). This condition
other resident species" such as sea ot-    reinterpretation of the faunal remains in   is further manifested both directly and
ters (Enhydra lutris) and their principal   Aleut middens to propose that (i) mul-    indirectly in the composition and stand-
prey. These interpretations presume that   tiple stable-state communities can be     ing crop of nearshore fishes. Islands
aboriginal man arrived in the New World    found historically and presently in the    dominated by sea otters characteristi-
as a "prudent predator" (3) and survived   Aleutian Archipelago and that (ii) aborig-  cally have high standing crops of species
as a wise manager of the natural re-     inal man, the Aleut in this case, was in-   that depend on and use sublittoral mac-
sources he exploited. These inter-      strumental in driving the community     roalgae for protection and spawning sub-
pretations also are consistent with the    from one stable state to another (Fig. 1).  strate. A characteristic detritus-based
popular hypothesis that paleoecologi-     To our knowledge this article is the first  food web supports most of these fishes
cal changes, such as Pleistocene extinc-   amalgamation of two theories, treating    through abundant populations of epiben-
tions of New World megafauna, were      aboriginal man as an important predator   thic crustaceans-mysids    and amphi-
caused directly by rapid environmental    through his influence on the nearshore    pods-which    are sustained by break-
change-climatic   and geological phe-    community.                  down of macroalgae (18, 19). In contrast,
nomena producing high rates of ex-                             islands without otters possess noticeably
tinction and speciation. There are, how-                          fewer nearshore fishes, and those pres-
ever, alternative hypotheses, such as     Alternate Communities            ent typically are species associated with
that proposed by Martin and Wright (4),                          the pelagic ecosystem and its food web.
positing that aboriginal man reduced or     Through intense predation, the sea ot-  This condition apparently has more far-
eliminated various large vertebrates up-   ter profoundly influences the organiza-   reaching effects on higher trophic forms,
on arriving in the New World. Results of   tion of nearshore communities in the     because islands without sea otters have a
recent ecological and archeological in-    North Pacific Ocean (7, 9, 10). We have   comparatively depauperate vertebrate
vestigations in the Aleutian Islands have   identified some of the more visible con-   fauna in terms of both number of species
prompted us to consider the Martin-      sequences of sea otter predation by com-   and abundance of individuals (7).
Wright hypothesis specifically for ab-    paring islands in the western Aleutian
original Aleuts.               Archipelago with and without sea otters     CharlesSimenstadis on the staffof FisheriesRe-
  Predation is important to the structure                         search Institute,College of Fisheries,Universityof
                       (7, 11-13). Differences between these    Washington,Seattle 98195;James Estes is a biolo-
and organization of many natural com-     two insular communities (Table 1) are    gist withthe AnchorageField Stationof the National
munities (5). The "keystone predator's"                          Fish and WildlifeLaboratory,U.S. Fish and Wild-
                       dramatic even to the casual observer.    life Service, and AffiliateAssistant Professor with
role (6) of sea otters is particularly dra-  Dense sea otter populations reduce her-   the Center for QuantitativeScience, University of
matic in that two alternate nearshore                           Washington;  Karl Kenyon is retiredfrom the U.S.
                       bivorous epibenthic macroinvertebrates    Fish and WildlifeService.
SCIENCE, VOL. 200, 28 APRIL 1978            0036-8075/78/0428-0403$02.00/0
                                          Copyright? 1978AAAS               403
  Since cessation of large-scale fur hunt-            in which diverse marine mammals, mac-        1969 (23). The strata and their major
ing in 1911, the sea otter has reestab-              roinvertebrates, and fishes were equally      faunal components are described in Fig.
lished its Aleutian populations through-              available for harvest. This interpretation     3 (26). These faunal data include the min-
out most of the archipelago, and in these             no doubt comes from investigations         imum number of sea otters and harbor
regions, the nearshore community is                showing that the Aleuts in the eastern       seals (Phoca vitulina) (27) and the gram
characterized by sparse populations of               Aleutians depended more on seasonally        dry weights of fish bones, sea urchin
sea urchins and abundant beds of macro-              abundant migratory food resources than       spines and tests, and limpet shells per
algae. Aboriginal Aleuts arrived in the              their neighbors did in the western Aleu-      centimeter of deposition. Minor com-
western Aleutian Islands about 2500                tians (24). Even these excavations, how-      ponents such as mussel and chiton shell,
years ago, although today they are ex-               ever, indicate disruption of the more sta-     and bones of northern fur seals (Callorhi-
tinct in that area.                        tionary component of the eastern Aleuts'      nus ursinus) and Steller's sea lions (Eu-
                                  food resources by overuse.             metopias jubata) (28), are not numerous
                                   The homogeneous composition and         enough to include graphically. Although
The Aleut                             stratified position of prominent faunal       the faunal remains are graphed by stra-
                                  components (Fig. 2) have suggested to us      tum, these strata are not discrete, equal
  Faunal remains in Aleut and pre-Aleut              another possibility, namely, that one or      time periods but are the archeologists'
kitchen middens excavated in the Aleu-               more shifts in the food subsistence base      designations of layers dominated by re-
tians probably are the best indication of             for aboriginal populations occurred dur-      mains of certain organisms-for     ex-
nearshore community structure during                ing Aleut occupation before intrusion of      ample, stratum E is the lens of sea urchin
prehistoric times (20-24). But Dall (20)              Western man. We have used the pre-         spines and tests seen in Fig. 2. The scale
and his successors have generally inter-              vious data and have reexamined faunal        is a measurement of depth from the sur-
preted stratified faunal midden remains              material from a prehistoric Aleut midden      face. Carbon dates at several depths in-
as different cultural periods, implying ex-            (49 Rat 31) (25) that was excavated on       dicate a uniform rate of deposition
ploitation of a single stable community              the Pacific coast of Amchitka Island in       (about 1 centimeter per 10 years), and




Tertiary             Harbor
consumers             seals




   dNr
Secondary         As        Deca-
consumers         oids       pods


                              i
             /    .        ..''


Primary            /  tons     Sea Ea
consumers s   Mussels      gastro-    rhinc s    c




         Nears  ore                             s  r
                                            shore

Producers                  Macroalgae       3        phyto-         alae         phyto-
         plankton                              plankton   /               Plankton




                                        I           .I..
Nutrients                            Algal                    Algal
  and                             detritus                   detritus

      404kto                                     VOL.o                       SCENE     200

           Sea otter-dominated                      Independent                  Sea urchin-dominated
Fig. 1. Generalized food web in the western Aleutian Islands emphasizing the effect of aboriginal Aleuts on the principal components of the
nearshore community. The sizes of circles indicate relative differences in standing crop between various components of the community in the two
alternate states of community organization. Arrows indicate the direction of biomass or energy flow; heavy arrows indicate importance or
magnitude of an interaction compared with the alternate community.
   404                                                                   SCIENCE, VOL. 200
therefore a moderately even time scale,    Aleuts induced a shift in the nearshore   3) suggests superficially that dominant
over the past 2500 years.           community toward an alternate structure   components of the community shifted
  Several assumptions are vital to inter-   as populations of invertebrates that were  from marine mammals and fish to her-
pretation of the midden faunal remains,    once limited by sea otters expanded with   bivorous invertebrates, then back to ma-
the most important being that these re-    the sea otter's decline. Many of the sea   rine mammals and fish. Since there is no
mains represent the availability of domi-   otter's principal prey are herbivores, and  obvious historical or biological ex-
nant food organisms for Aleut harvest     these populations probably grew be-     planation for the second shift, these data
from the nearshore community. Overall,     cause of an abundance of algae and the    might be interpreted as evidence for cul-
skeletal, calcareous, and shell remains of   release from intense predation. These    tural changes in Aleut use of temporally
food organisms are well preserved. The     herbivores invaded the sublittoral fringe  uniform food resources, or perhaps as
most notable exception is the lack of     and littoral zones where they became     cyclic overuse of both marine mammals
bones from Pacific salmon (Oncor-       available for harvest by Aleuts. Even-    and herbivorous invertebrates. In the ab-
hynchus spp.), Dolly Varden (Salvelinus    tually an alternate state of community    sence of additional information, a con-
malma), and the smooth lumpsucker       organization was attained.          vincing argument could not be made for
(Aptocyclus ventricosus), which we                             these, or perhaps other, alternative ex-
know to have been harvested commonly                            planations.
throughout the Aleutians (20, 24, 29).     Sea Urchin Size Frequencies           Comparison of the size and distribu-
These fish bones and macrocrustacean                            tion of sea urchins between prehistoric
exoskeletons apparently were too fragile     Stratigraphic variation in the abun-   and present communities clarifies the sit-
or not calcified enough to be preserved.    dance of midden faunal remains, togeth-   uation. The size of sea urchin remains in
Soft-bodied mollusks such as cephalo-     er with prehistoric and present size class  middens cannot be measured directly,
pods are not represented in the faunal re-   distributions of sea urchins, provides    owing to fragmentation of the tests; how-
mains for a similar reason. We further     evidence for the general pattern of spa-   ever, a conspicuous feature of these re-
assume that the Aleut harvested food in    tial and temporal changes in nearshore    mains at Amchitka is the large size of
proportion to availability, so that major   community organization at Amchitka Is-    calcareous parts of the oral apparatus, or
shifts in harvesting strategies were im-    land during the past 2500 years. The in-   Aristotle's Lantern, which are found in-
posed by changes in the availability of    verse relationship between abundance of   tact amid the broken tests and spines.
harvestable organisms. Corresponding-     sea otters and grazing invertebrates (Fig.  Strongylocentrotus polyacanthus is the
ly, we assume that the remains of dif-
ferent food organisms were not dis-      Table 1. Comparativestatusof nearshorecommunitiesin Rat Islandsand Near Islands,western
carded in different areas, and that faunal   AleutianIslands, Alaska.
remains in the vertical profiles through
                           Species         Rat Islands            Near Islands
the middens represent changes in the
                          (sources)      (AmchitkaIsland)(48)       (ShemyaandAttu islands)
composition of food exploited by the
Aleuts through time (30).           Sea otters      Abundantfor at least last several  Sparse;first sightingin late 1960's
                         (31,49)       decades; current estimated     after exterminationby fur tra-
  The data in Fig. 3 indicate a strong
                                   populationgreaterthan 6000     ders; currentpopulationon At-
negative relationship between the har-                                  tu about 350; none at Shemya
vest of sea otters, fish, and harbor seals,
                        Macroalgae      Abundant; diverse epibenthic    Rare; restricted to a few species
on the one hand, and the harvest of sea     (7, 10)        canopy(principally four species   isolated in sublittoral
                                                                fringeand
urchins and limpets on the other. We in-               of Laminaria, Agarum cribro-    sublittoralpatches
terpret this as evidence that (i) the avail-             sum, Rhodophyta spp.) and a
                                   dense surface canopy (Alaria
ability of prey items preferred by Aleuts
                                   fistulosa); competitive inter-
changed greatly during the time Aleuts                actions predominate
occupied Amchitka and (ii) this change     Sea urchins,     Rare; maximum test diameter     Dense; maximum test diameter
was caused largely by Aleuts over-        S. polyacanthus   < 32 mm; increasing density    > 100 mm; highest density and
harvesting or harassing sea otters, with     (7)         and size with depth        greatest individualsize at sub-
the consequence that during at least the                                 littoralfringe
past 2500 years the nearshore commu-      Limpets(11),     Density 8 m-2 and maximum      Density 82 to 356 m-2 and maxi-
nity at Amchitka shifted between one       C. pelta (50)    length51 mm             mum length 67 mm
dominated by sea otters to one charac-     Chitons,K. tuni-   Rare;density < 1 m-2        Common;density 32 m-2
terized by few sea otters and an abun-     cata, C. stel-
dance of large invertebrate herbivores.     leri (7)
That the Aleut was technically capable     Mussels, Mytilus   Rare and small;density 3.8 m-2   Common and large; density 722
of locally reducing or eliminating sea ot-   edulis, Modio-                      m-2
                        lus sp. (7)
ters during prehistoric times is supported
by the near elimination of sea otters from   Barnacles,Balanus  Rare and small;density 4.9 m-2   Common and large; density 1215
the North Pacific Ocean after the en-      glandula,B.                       m-2, dominating upper littoral
                         cariosus (7)                       zone
slavement of Aleut hunters by Russian
                        Nearshorefish    Abundant, diverse fauna; high    Sparse fauna outside littoralzone
fur traders (31, 32).
                        (18,51)       standingcrop supportedby al-     except for deepwaterdemersal
  The effect of Aleut exploitation was               gae detritus-basedfood web      and neritic forms and popu-
therefore twofold: (i) By overexploiting                                 lations associated with sparse,
sea otters, Aleuts limited the availability                                isolated patches of macroalgae
of this prey, forcing a change in harvest-   Harborseal,     Estimated density, 8.1 per kilo-  Estimated density 1.5 to 2.1 per
ing strategy to increasingly more avail-    P. vitulina(52)    meter of coastline; frequently   kilometer of coastline; seldom
able organisms such as sea urchins and                observed in groups larger than    observed in groups larger than
                                   50 animals              ten animals
limpets; and (ii) in limiting the sea otter,
28 APRIL1978                                                           405
                   only species of sea urchin known to have
     Fig. 2. Stratigraphic   inhabited the western Aleutians during
     profile (front face of
     excavation units) of    the Recent epoch. Therefore we suspect-
     midden site 49 Rat     ed that some parts of Aristotle's Lantern
     31, Amchitka Is-      in the midden remains might indicate the
     land. *, 14C sample    size of Aleut-harvested urchins. The
     locations. The fig-
     ure (with minor rela-   demipyramids, which are the thickest
     beling) and the fol-
                   and most robust parts of Aristotle's Lan-
    = lowing legend are      tern, were chosen as the most likely in-
   D reproduced      from  dicator, since they are not prone to wear
     Desautels et al. (23,   and regeneration from grazing.
     figure 15): A, Dark
     brown highly organ-      The correlation between sea urchin di-
     ic  humus    (root  ameter and demipyramid length was de-
     zone), mixed with     termined from living specimens collected
     sand and overriding    at Amchitka and Shemya Islands (Table
     a thin lens of dark
               Dark  2). We found that linear regressions of
     clay.   B,
     brown-colored mid-     test diameter and demipyramid length
     den and sea urchin     were not significantly different between
     mixture with a low     Amchitka and Shemya (F2,212 3.61, =
     and sporadic con-     P > .05), and that the common regres-
     tent of sand and
     black clay. C, Light    sion function
     brown sea urchin
     and midden mixture         yi = -5.9484 + 5.1732 xi
     with   light-colored
    - sea urchin lenses     where yi equals test diameter and xi
     and sand. D, Dark     equals demipyramid length, was ex-
     brown sea urchin      tremely precise (r = .9838).
     and midden mixture      The high correlation between urchin
     with a high concen-
     tration of inter-
                   diameter and demipyramid length has al-
     mixed sand and       lowed us to estimate accurately and pre-
     clay; constitutes a    cisely the size of sea urchins harvested
     basement layer of     by Aleuts at Amchitka. Figure 4 illus-
     C. E, Light yellow-    trates size frequency histograms by stra-
     ish-white    colored
     sea urchin and lim-    tum for sea urchins deposited in the Am-
     pet mixture. F, Pure    chitka midden, together with comparable
     fish bone. G, Dark     data from recent collections from the lit-
     brown sea urchin,     toral and shallow sublittoral zones at
     sand, and midden      Amchitka (12) and Attu Islands (15).
     mixture. K, Light
     brown sea urchin      These data demonstrate that the size-fre-
     and midden mixture     quency distributions of sea urchins gath-
     with deposits of      ered by Aleuts occupying the midden
     pure sea urchin; be-    were virtually constant throughout the
     comes    discolored
   C and in
          spots more    period of Aleut occupancy. Only "M
     compact    towards   stratum," representing the earliest peri-
     the base of the stra-   od of occupation of this site, provides no
     tum. H, Dark black     record of sea urchins. Most important,
     greasy, highly or-     these size-frequency distributions typify
     ganic, with a high
     concentration     of  present-day communities devoid of sea
     fish bone. I, Dark     otters, as shown by the data from Casco
     brown sea urchin      Point [see also (7)] which is outside the
     and midden mixture
     with an increased
                   range of the small population of sea ot-
     amount of sand;      ters now occupying that island (33). In
     contains thin inter-    contrast, these distributions contain
     mixed lenses of (i)    larger sea urchins than we found either at
     black organic mate-    Pisa Point (34), which now is in the cen-
     rials and (ii) light
                   ter of the sea otters' range on Attu, or at
     sea urchin lenses.
     J,  Pure    brown   Amchitka where sea otters are currently
   >  sand; scattered fish    abundant (35). Furthermore, whereas
   ._ and mammal bone.      sea otters have been abundant at Am-
   C  M, Yellowish-brown     chitka for at least several decades, in
     sand; oxide layer
                   contrast with the small, recently estab-
     appearing at base.
     S, Indicates sterile;   lished population at Attu, the size-fre-
     starting with a pure    quency distributions of sea urchins at
     clay lens and contin-   Pisa Point on Attu and those from Am-
     uing down into a      chitka are nearly identical. From these
     light brownish-gray
     sand layer.        observations and data we conclude that
                   even a sea otter population at low den-
406                             SCIENCE, VOL. 200
sity rather quickly causes a noticeable    Table2. Sources of sea urchin,S. polyacanthus, used in correlationbetween urchindimension
shift in the size-frequency distribution of  and demipyramid (componentof Aristotle's Lantern,mouth structure)size.
sea urchins toward smaller individuals.                                     Depth        Sample         Urchintest
  The reconstructed size-class distribu-     Location        Collectionmethod           range        size          diameter
tions of sea urchins (Fig. 4) therefore im-                                    (m)          (N)         range (mm)
ply that a community lacking or nearly     AmchitkaIsland       Bottomtrawl             82 to 92          58         32 to 80
devoid of sea otters persisted (at least lo-
                        AmchitkaIsland       Scubacollection            6 to 31          62         5 to 52
cally) throughout the time Aleuts occu-
                        ShemyaIsland        Scubacollection            3 to 23          96          6 to 74
pied Amchitka. Aleuts probably selec-
tively gathered the largest urchins avail-
able to them, and although such selective
behavior would tend to mask minor       benthic fishes, and several seasonal or           occupy the nearshore community as a
changes in the size-frequency distribu-    transient inhabitants of nearshore com-           nursery area. In the eastern Aleutians,
tion of sea urchins over time, it could not  munities such as Atka mackerel (Pleuro-           where these fishes are generally more
account for the distributions observed in
                        grammos monopterygius), Pacific hali-            abundant, they constituted more signifi-
the midden strata if many sea otters were   but (Hippoglossus stenolepis), and rock           cant food resources and contributed to
present (36).                 sole (Lepidopsetta bilineata). These spe-          seasonal patterns in resource exploi-
  The most reasonable interpretation of    cies are not directly dependent on near-          tation by the Aleuts of that region (24).
midden faunal remains is that there was    shore communities for food or pro-              The second component of the fish
some spatial disparity in Aleut hunting    tection, although they may use these wa-          fauna includes species that are more per-
and gathering activities. We suggest that   ters periodically for spawning and, as           manent members of the nearshore fish
Aleuts gathered sea urchins and limpets    with the rock sole, their juveniles may           assemblage, including rock greenling
near the villages-areas from which sea
otters were harvested or harassed to
near extinction. Later hunting (and per-
                                                              Sea otters
haps fishing) activity was apparently di-                           2.0-

rected toward more distant areas, per-
                                         X      1.6-
haps even other islands. This ex-
planation is most plausible because even
                                         4-

                                           n    1.2-
sparse populations of sea otters cannot                      o

                                         E      0.8-
occur in the same place as sea urchins of
the size gathered by Aleuts (37).                        0
                                         z      0.4-
                                         E
                                         E       0-
The Fish Assemblages                               E      0.8-             Harbor seals

                                               04-
  Abundance of fish in the various mid-                 E
den strata is correlated with the abun-                          -0-                          I             ,
dance of sea otters (Fig. 3). This pattern                                         Sea urchins
follows logically from our recent findings                          40-
that the abundance of nearshore fishes is                           30-
positively correlated with the abundance
of macroalgae, and therefore with a high-                           20
                        Fig. 3. Principal fau-
density population of sea otters (38).     nal remains in the
                                                10-
However, although the relationship be-     strata of midden 49
tween aboriginal Aleuts, sea otters, and    Rat 31 at Amchitka             O     '
                                         0
                        Island. The strata
certain herbivorous macroinvertebrates                                             Limpets
                        were designated by
seems fairly clear, the interpretation of                    40.
                        Desautels et al. (23).
coincident availability and harvest of                      E
                                         {      30-
specific nearshore fishes is more com-
plicated. The relative abundance of prin-                           20-
cipal fish species occurring in the midden
                                         1      10-
strata is illustrated in Fig. 5. These data                          0-
were derived from estimates of the mini-                     o       o
mum numbers of fish, based on the abun-                                             Fish
dance of characteristic head bones (39).                           40-
  Information concerning Amchitka's
                                               30_
recent fish communities (18) suggests
that the marine fish assemblage available                           20
to the Aleuts included two components,
                                                10-
only one of which was directly tied to the
structure of the nearshore community.                                                 ,
One component includes species prob-                  Deposition (crn)            6      i    1E 140 !       Ig ,
                                                                              1
                                    Strata                  I           E      ID|  C   |B[A
ably little affected by Aleut fishing pres-
                                    Carbon date                                     1080
sure or by kelp abundance, such as off-                                 95BC    95B                  +80 AD
shore (> 40 m depth) demersal or epi-                                        Midden stratigra phy
28 APRIL 1978                                                                          407
(Hexagrammos lagocephalus), red Irish which, although also found in deeper wa-                                   substrate, or indirectly on the detritus-
lord (Hemilepidotus   hemilepidotus), ters offshore, occupies the nearshore wa-                                 based food web, they represent popu-
rockfish (Sebastes spp.) (40), great scul- ters during much of the year. These spe-                                lations which (i) could have been over-
pin (Myoxocephalus polyacanthocepha-    cies characterize the otter-dominated                                  exploited and (ii) should have been re-
lus), and smooth lumpsucker (A. ventri- community at Amchitka, or once did (18,                                  duced with expansion of the sea urchin
cosus). Pacific cod (Gadus macrocepha- 31). By their reliance directly on the kelp                                 population and declining kelp abun-
lus) represents a transitional species community for protection and spawning                                    dance.
                                                                           Apparently the Aleut, by controlling
                                                                          the abundance of sea otters, indirectly
    40                                        Strata                          influenced the concurrent abundance
                                              J                           (Fig. 3) of these fishes. Data from the
    30-
                                             n--3                           midden strata (Fig. 5), in conjunction
    20                                      d    .69.3 mm
                                                                          with our recent collections at Amchitka
    10
    n                    II                                                and Attu, support this conclusion. Fishes
                                                                          of the exposed, rocky nearshore habitat
                                                                          were more abundant at Amchitka than at
                                                I                         Attu (as much as 44 times the catch per
                                             n -23
                                               62.5 mm
                                                                          unit effort), although percentage compo-
                                                                          sition of species was not strikingly dis-
                                                                          similar. Rock greenling predominated in
                                                                          both communities and, when the small
                                               H                          patches of kelp bed habitat persisting at
                                             n=5                            Attu were sampled, catch per unit effort
                                              =
                                              -63.6 mm        C
                                                                          for this species was similar to that of
                                                         -0
                                                         -0                Amchitka. Thus we believe that the
                                           -              E                availability of nearshore fishes is strong-
                                                                          ly correlated with the abundance of
                                                                          macroalgae.
                                           c/= 70.6 mm
                                                                           Nearshore fish species (rock green-
                                                         m
                                                         X
                                                         _.                ling, red Irish lord, and Pacific cod) typi-
                                                         e-                cally were exploited more successfully
                                                         E   Fig. 4. Sea urchin     than offshore species (Fig. 5). While the
                                               D             size-frequency distri-   abundance of both components is corre-
                                             n=75              butionsfrom strataof    lated with patterns of sea otter/urchin
                                           dc 61.6 mm             midden 49 Rat 31 at    abundance (Fig. 3), the nearshore com-
e                                            2
                                                            Amchitka, and from
                                                            present-day commu-     ponent   consistently   predominates
u
c
4)1
                                                            nities at Amchitka     throughout all strata. This suggests that
                                                            and Attu islands. An    Aleut fishing was directed principally at
)1-)
                                                            explanation of loca-    nearshore areas and that offshore species
                                                            tions is given in the
LL-                                         c X679     mm
                                                            text. Strata levels cor-  were probably caught incidentally to
                                                            respond with those     the nearshore component. The pre-
                                                            given in Fig. 3. Abbre-  dominance of nearshore fish remains in
                                                            viations: dj = mean    the midden also supports the argument
                                               B             diameter; n = sample    that fluctuations in fish abundance (Figs.
                                             n=14              size.
                                                                          3 and 5) were an effect of overexploita-
                                           c/-65.9 mm
                                           2                              tion of sea otters by Aleuts and the con-
                                                                          sequences to the nearshore community.

                                             CASCO PT, AT TU (-3 m)
                                             n-129
                                                                          Harbor Seals
                                           cd =61.7mm
                                           2
                                                                           The distribution of harbor seal bones
                                                                          through the midden strata suggests a pat-
                                                                          tern of availability and exploitation simi-
                                             PISA PT, ATTU(-3 m)                    lar to that of the sea otter (Fig. 3). Har-
                                             n =875
                                           cd =22 9 mm
                                                                          bor seals may have been harvested op-
                                           2
                                                                          portunistically during periods when
                                                                          Aleuts hunted marine mammals. If ma-
                                                                          rine mammal hunting was more intense
                                             AMCHITKA( nter tdal)
                                                                          during those prehistoric periods when
                                             n-224           (c               sea otters were abundant, then the ob-
                                           cd =18.9mm
                                           2                              served pattern of use of harbor seals
                                                          E
                                                          c:
                                                                          would be expected, even if the abun-
               ?                T,
                                                                          dance of seals remained nearly constant.
    V ,  I,,  ,  ,  ,I  I  ,  ,      ,      ,  I,  I,  ,
     o
                       I  I    I


        10 20 30 40 50 60 70 80 90 100 110                                                 Harbor seals probably are closely linked
          Sea urchin      diameter         (mm)                                      with the nearshore detritus-based food
    408                                                                                 SCIENCE, VOL. 200
web throughtheir consumptionof near-                     2.5
                                          Nearshore: o Gadus macrocephalus
shore fishes (41). Therefore, a relatively                           A Hexagrammos lagocephalus
                                                O Hemilepidotus hemilepidotus
high abundanceof harborseals is a pre-                             0 Sebastes sp.
dictableconsequence of abundantsea ot-               -     2.0  Offshore:  *  Pleurogrammus monopterygius
ters in the community,andthis increased               E              * Pleuronectidae sp.

availabilityof harborseals wouldexplain   Fig. 5. Abundance   c
                                  ax

theirincreaseduse duringtimes when, or    (minimumnumber     u

                       per centimeterof    a   L1.5
in areas where, sea otters were abun-    strata)of dominant
dant. There is some supportfor this hy-   marinefish species   -.-


pothesis from our observation that har-   in strataof midden   0
                                   C
                       49 Rat 31. Strata    a,
bor seals appearto be more abundanton    levels correspond   E
                                     -
                                       1.0
Amchitkathan on Attu (Table 1) (7).     with those given in  c
                                  E
                       Fig. 3; no fish data  E
                       wereavailable from
                       D stratum.       i     0.5
Discussion

  Naturalcommunitiescan exist at mul-
                                        0  ---  =J
tiple stable points in space or time (8)-a                              I    H     E    C
stable point being characterized by a                                      Midden strata
specific structuraland functionalassem-
blage of species in a communitywhich is
persistentthroughtime and recognizably    sources. Therefore, the presence or ab-         Pleistocene (45), and they were common
different from other assemblages that    sence of sea otters in the nearshorecom-         in the CommanderIslands until shortly
can occur in the same space. This defini-  munity is a driving force toward either         afterG. W. Steller firstobservedthem in
tion charges us to examine communities    one of two alternatestable points.            1741 (46). They apparently fed on the
and to interpretcommunitychanges with      We envision that evolution in the          surfacecanopy(47), and theirrole as her-
appropriate  referenceto time and space.   western Aleutian nearshore community          bivores in the nearshorecommunitywas
Because several important predatory      proceeded under a suite of selective          no doubt an importantone.
species in the western Aleutian Islands    forces which were associated closely           Despite these uncertainties, it is evi-
are highly motile (for example, Aleuts    with the presence of sea otters as a key-        dent that the arrivalof the Aleut served
and sea otters), the appropriatespace     stone predator.Most of the largerAleu-         as a driving force toward the alternate
may be as large as islands or island     tian Islands were extensively glaciated         stable communitystate by effectively re-
groups. The appropriate time may be      during the Pleistocene (43). Precursors         movingthe sea otter as a keystone pred-
decades or centuries, consideringthe life   to the contemporarycommunitiesin this          ator and replacing it at a higher trophic
histories of the communities' "founda-    area probablyexisted in refuges associ-         level. Indications are that this change
tion species" (42) such as Aleuts, sea ot-  ated with the Asian and North American         dramaticallyeffected a new structure,
ters, sea urchins, and various perennial   continents where they persisted and           composition, and organization in the
brown algae. Indeed, the communities     evolved with the predecessors of mod-          nearshorecommunity.
described in this article have been suffi-  ern-daysea otters (Enhydra)since about
ciently persistentthroughtime and space   the Pliocene (44).
so that there can be little doubt they are   The community probably evolved to-          Conclusion
locally stable in this context.        warda relatively stable state in the sense
  The question thus becomes, Why is a    that it apparentlywas resilient to minor          Contrary to popular opinion, it is
particularstable state observed at a par-  perturbations that it did not undergo
                              and                    likely that aboriginal mandirectlycaused
ticular point in time and space? Suther-   majoroscillationsthroughtime. We base          the extinction of certain New World
land(8) arguedthat the explanationoften   this conclusion on the high longevity of         megafauna during the Pleistocene (4).
is found throughexaminationof specific    many of the foundation species in the          Evidence for this conclusion generally
historicalevents and the consequent un-   present-day community, together with           has been in the form of temporal-spatial
derstanding of how these events may     the observationthat populationsof these         correlations between the extinction of
have led to the presence or absence of    species are not known to fluctuategreat-         species and arrivalof aboriginalman. In
key consumers in the community. His-     ly under natural circumstances. Selec-          this article we have employed a some-
tory in this instance has provided us in-  tive forces controlling the evolution of         what different approach by treating ab-
sight into the relationshipbetween the    these patterns apparentlywere centered          originalAleuts as key predatorsand as-
arrivalof aboriginalman to the Aleutian   on the control of herbivoresby sea otters        sumingthat, as such, their activities are
Islands and the initiationof shifts in the  and the consequent development of a           revealed by characteristicbiotic assem-
structureof the nearshore marine com-    macroalgalassociation that served as a          blages that can be interpreted the light
                                                                   in
munity to alternate stable states. The    requisiteresource to many other species         of a contemporary understanding'of
mechanismfor this change is the remov-    of animalsin the community.               communitydynamics.
al of a keystone predator,which, by defi-    As Dayton (10) pointed out, such hy-           The ecological interaction critical to
nition, preferentiallyfeeds on prey that   potheticalspeculationconcerningevolu-          our interpretation of the activities of
are capable of excluding subordinate     tionary adaptation is frequently com-          aboriginal Aleuts is that dense popu-
species throughcompetitionfor a requi-    plicated by unknown interactions in-           lations of sea otters in the western Aleu-
site resource such as food or space. The   volving recently extinct species-in this         tian Islands limit sea urchins to sparse
sea otter is clearly such a predator:its   case Steller's sea cow (Hydrodamalis           populationsof small individuals.In turn,
foraging activities prevent sea urchins   gigas). Sea cows are known to have in-          this interactionis importantto the main-
from dominating food and space re-      habited Amchitka Island until the            tenance of robust kelp beds and a rich
28 APRIL 1978                                                             409
associated fauna of fish, birds, and ma-        15. J. A. Estes and C. A. Simenstad,unpublished        the northeastend of the islandandis highlypro-
                              data gatheredduringthe summersof 1976 and         ductive.Duringsummer1976we saw 344 sea ot-
rine mammals. Midden remains suggest            1977at Attu Island.                    ters in the course of surveys, 33 percent of
that aboriginal Aleuts locally disturbed        16. 0. I. Koblents-Mishke,  Oceanology(USSR) 5,       which were pups still associatedwith females.
                              104(1965);B. W. McAlister,BioScience21, 646      34. Datafromsamplescollectedat only the 0 to 3-m
this system by overexploiting the sea ot-          (1971); C. C. Amundsen and E. E. Clebsch,         depth during summer 1976 are presented here
                              ibid., p. 619; K. H. Mann,Mar. Biol. 14, 199        because this depth range probablyrepresents
ter, thus minimizing or eliminating its           (1972);Science 182, 975 (1973).              the area-the littoral bench and sublittoral
keystone maintenance role in the com-          17. Althoughthe nutritional  requirements macro-
                                                   of        fringe-from which aboriginalAleuts gathered
                              herbivores (particularly  urchins)are presently                      in
                                                            sea urchins.Some variation the size frequen-
munity. Consistent with predictions             unknown,populationsat areas without sea ot-        cy distribution sea urchinsis relatedto such
                                                                     of
based on observations of communities            ters are characterized high biomassand low
                                          by                 factors as exposure and depth; however, this
                              productivity.Preliminary  resultsof our ongoing     variationis small comparedwith the magnitude
with and without sea otters, the abun-           studyat Attu suggestthatsea urchinpopulations       of the differencedepictedbetween Casco Point
dance of sea otter bones through the            maintain themselves(i) by congregating  nearthe     and Pisa Pointin Fig. 4, whicharegenerallyrep-
                              sublittoral fringewherebythey exploitalgaeand       resentative of those two areas. Sea urchins
midden strata is directly related to the          detritus washed from the robust algae assem-        about equal in size to those we observed at
abundance of marine fish and seals, and           blageof the littoralzones; (ii) by fastingor con-     ShemyaandAttuwere collectedfrom90- to 100-
                              sumingfoods of lower nutritional  value, such as     m depth at Amchitkain a bottomtrawl by the
inversely related to the abundance of sea          corallinealgae, diatoms,or animaldetritus;and       M.V. Commander(University of Washington
                              (iii) throughallocating most of their nutritional     charterresearchvessel). There is no evidence,
urchins and limpets.                    inputto maintenance way of a naturally
                                         by           slow     however, to suggest that Aleuts harvestedsea
  Specific life history adaptations and in-        growthrate (examination growthringson the
                                            of               urchinsfrom these depths.
                              interambulacral plates of urchinsfrom Attu and                    at
                                                          35. The sea otter population Amchitka    Islandcur-
teractions among species in this commu-           Shemyaindicatethat individuals   commonlylive      rentlyis nearcarrying   capacityandwas recently
                              to ages beyond20 years)and reducedreproduc-        estimatedto containmorethan6000animals(11,
nity probably evolved, to a large extent,          tive effort (suggestedby small gonads and low       49).
either directly or indirectly in response          recruitment small size classes).
                                     to                    36. Ottersand large sea urchinscould coexist in a
                            18. C. A. Simenstad,J. S. Isakson,R. E. Nakatani,       more intimatespatialassociationif there were
to the keystone disturbance role of sea           in TheEnvironment Amchitka
                                         of      Island,Alaska,     refuges,such as substratecrevices, in whichur-
otters. This role probably was constant           M. L. Merritt and R. G. Fuller, Eds. (TID-         chinscouldescape predation otters(9). How-
                                                                            by
                              26712, Energy Researchand DevelopmentAd-          ever, the rocky sublittoral  substrate the west-
                                                                                of
and persistent over relatively long time          ministration, Oak Ridge, Tenn., 1977),p. 451.       ern Aleutian Islands is essentially flat or of
periods because sea otter populations          19. Althoughquantitative  dataare still lacking,kelp     broadrelief in all areas that we have examined.
                              bed communitiesof the northPacificappearto         This substrateis covered by a smooth,continu-
probably were seldom, if ever, subjected          be analogousto eelgrass communitiesin their        ously flatpavementof encrusting   corallinealgae
to disruptive disturbances from pre-            supportof a food web based on detritusand as-       (Clathromorphum    spp.). Apparentlythe only
                              sociatedmicrofauna.                    refuge  sea urchins have from predationby sea
dation or climatic-geological catastro-         20. W. H. Dall, On the Succession in the Shell-        otters is in deep (> 100 m) offshore water
                              Heaps of the AleutianIslands, Contributions   to    beyond the otter's effectivedivingdepth.
phes. For these reasons we conclude             North AmericanEthnology(Government      Print-  37. Besides the Aleut preclusion sea otters, three
                                                                            of
that the nearshore community had little           ing Office,Washington,  D.C., 1877).           alternativehypotheses, stated below, might be
                            21. W. Jochelson,ArchaeologicalInvestigationsin        advancedto explainthe patterns faunaldistri-
                                                                              of
inertia against predation of sea otters by         the AleutianIslands (Publ. 367, CarnegieInsti-       butionthrough middenstratathatare seen in
                                                                     the
aboriginal Aleuts. Changes in the com-           tutionof Washington,  Washington,  D.C., 1925).     Fig.  3. Althougheach of these may be to some
                            22. V. E. Ransom,Am. Anthropol.48, 607 (1946);         extent true, we have rejectedthem as principal
munity that followed this disturbance            W. S. Laughlin W. G. Reeder,Science 137,
                                       and                    explanations the reasonsstatedwith the hy-
                                                                    for
                              856 (1962);J. P. Cook, E. J. Dixon, Jr., C. E.       potheses:
consequently were for the most part dra-          Holmes, Holmes and NarverRep. HN-20-1045            1) Fluctuationsin faunalabundancethrough
matic and not preadapted for.                (Las Vegas, Nev., 1972).                  the midden reflect culturalchanges by Aleuts.
                            23. R. Desautels, USAECRes. Dev. Rep. AT(29-2)-        Perhapsthe strongestsupportfor this hypothe-
        References Notes
              and               20 (1970).                         sis lies in the second peak in otter abundance
                            24. C. C. TurnerII, L. R. Richards,J. A. Turner,        (strataC and B, Fig. 3). Although Aleutcul-
                                                                               the
1. J. M. Hett and R. V. O'Neill,Arct. Anthropol.      paper presentedat XLI International    Congress     ture certainly must have changed during the
  11, 31 (1974);A. P. McCartney,in Prehistoric      of Americanists, Mexico City, 2 to 7 September      2500years that they occupiedthis AmchitkaIs-
  Maritime Adaptationsof the Circumpolar  Zone,     1974.                           land midden,we reject this as an alternative   to
  W. Fitzhugh,Ed. (Mouton,The Hague, 1975),      25. The middensite 49 Rat 31 is located on the Pa-       the otter preclusionhypothesison the basis of
  p. 181;W. S. Laughlin J. S. Aigner,in ibid.,
              and               cific Ocean side of AmchitkaIsland on a bluff       sea urchinsize class distributions  shown in Fig.
  p. 281;A. P. McCartney, TheEnvironment
               in         of    frontinga small bay. A large streamdrainsinto       4. If the culturalchangehypothesisis correct,it
  AmchitkaIsland, Alaska, M. L. Merrittand R.       the bay adjacentto the middensite and a broad       would predict(as the result of sea otters being
  G. Fuller, Eds. (TID-26712,Energy Research       expanse of intertidalbench characterizesthe        abundantin the community)smallersized sea
  and Development Administration,   Oak Ridge,     coastlineto the east. Extensivekelp beds pres-       urchinsthan we observed throughthe midden
  Tenn., 1977),p. 59.                   ently exist immediatelyoffshore. A more de-        strata.
2. W. S. Laughlin,Science 189, 507 (1975).         tailed descriptionof the site and its excavation        2) Fluctuationsin faunalabundancethrough
3. L. B. Slobodkin,Am. Zool. 8, 43 (1968).         and artifactsis given in (23).               the middensreflectnaturalchangesin the com-
4. P. S. Martin H. E. Wright, Eds. Pleisto-
         and        Jr.,         26. Faunal remains were sifted in 1/8-inchscreen        munity.We rejectedthis hypothesisfor two rea-
  cene Extinctions;The Searchfor a Cause (Yale      sieves (23).                        sons. There is no evidence that sea otter popu-
  Univ. Press, New Haven, Conn., 1967.)        27. Identificationand extrapolationof minimum         lations fluctuategreatly under naturalcircum-
5. J. H. Connell,Ecol. Monogr.31, 61 (1961);J. L.     number sea otter andseal remains stratum
                                   of               per         stances. Even if otters did fluctuatein abun-
  Brooks and S. I. Dodson, Science 150, 28        were originallymadeby J. S. Aigner,University       dancefrom time to time, it is difficult believe
                                                                                 to
  (1965);R. T. paine,Am. Nat. 100, 65 (1966);J.      of Connecticut,Storrs.                   thatthe population   wouldhave declinedto such
   L. Harper,in Diversityand Stabilityin Ecologi-   28. The paucityof sea lion bones in 49 Rat 31 is sur-     low numbersand remained    depressedfor nearly
  cal Systems, BrookhavenSymp. Biol. 22 (Na-       prisingin view of the abundant lion remains
                                               sea           2500 years as the data in Fig. 4 would predict
  tional Technical InformationService, Spring-      reportedfrom middensin the eastern Aleutian        underthis hypothesis.
  field, Va., 1969);D. H. Janzen,Am. Nat. 104,      Islands. Presently, there are relativelyfew sea        3) Aboriginal  Aleuts overexploitedotters but
  501 (1970);ibid. 110, 371 (1976);Annu. Rev.       lions at Amchitkadespitethe fact thatthey have       replacedthe ecologicalroleof ottersas keystone
  Ecol. Syst. 2, 465 (1971);ibid. 7, 347 (1976);P.    not been exploitedfor years, and it may simply       predators  in the nearshorecommunity.   This hy-
  K. Dayton,Ecol. Monogr.41, 351 (1971);J. W.       be that there never were manysea lions at Am-       pothesisis rejectedbased on the sea urchinsize
  Porter,Am. Nat. 106, 487 (1972);P. K. Dayton,      chitkafor Aleuts to harvest.                frequencydistributionsin Fig. 4. If Aleuts re-
  G. A. Robilliard,R. T. Paine, L. B. Dayton,     29. W. Jochelson, History, Ethnologyand Anthro-         placedthe otters' ecologicalrole, we wouldpre-
  Ecol. Monogr. 44, 105(1974).              pology of the Aleut (Publ.432, Carnegie   Institu-    dict an absenceof largersize frequencydistribu-
6. R. T. Paine,Am. Nat. 103, 91 (1969).          tion of Washington,Washington,   D.C., 1933).      tions of sea urchinsthroughthe midden strata
7. J. A. Estes and J. F. Palmisano,Science 185,    30. Thisassumption supported evidencein (23,
                                       is      by              than we found.
   1058(1974).                      pp. 36-37) showingthat the verticaldistribution    38. This interpretation  assumesthat fishingwas not
8. J. P. Sutherland, Am. Nat. 108, 859 (1974).       of stratais roughlyuniformhorizontally in  and      restrictedto those areas from which sea otters
9. L. F. LowryandJ. S. Pearse,Mar.Biol. 23, 213      consistent temporalsequence throughthe pro-         were eliminated. If fishing activities were as
  (1973);J. Cooper, M. Wieland,A. Hines, Verli-      fileof midden49 Rat31. Severalminorstrataare        widespread marinemammal
                                                                   as         hunting, then the
  ger 20, 163(1977).                    horizontallydiscontinuous,hence the omission        ecological model presentedin this article pre-
10. P. K. Dayton, U.S. Fish Wildl.Serv. Fish. Bull.     of strataF, G, K, and L fromFig. 3, and D, F.       dicts thataccess to sea ottersis positivelycorre-
  73, 230 (1975).                     and G from Fig. 5.                     lated with access to nearshorefish.
11. J. A. Estes, thesis, University of Arizona     31. K. W. Kenyon,TheSea Otterin the EasternPa-       39. The notableexceptionis the surprising    absence
  (1974).                         cific Ocean (Government   Printing Office,Wash-      of readily identifiablehead bones from Pacific
12. J. F. Palmisano,thesis, Universityof Washing-      ington,D.C., 1969).                    halibut.Some of the numerousvertebrae,how-
  ton (1975).                     32. The hunting  technologyof Aleutsunderenslave-       ever, may have belongedto this species. The lit-
13. _    and J. A. Estes, in TheEnvironment  of     ment by Russian fur traders remained un-          erature documents this fish as a major food
  AmchitkaIsland, Alaska, M. L. Merrittand R.       changed from aboriginalmethods, except that        source of the Aleut in historic times (21, 29).
  G. Fuller, Eds. (TID-26712,Energy Research        largeships were used to transport  Aleuts to the     This suggeststhat either(i) the halibutwas har-
  and Development Administration,   Oak Ridge,      most favorablehuntinggrounds[W. R. Hunt,          vested in prehistoric  periodsbutwas not treated
  Tenn., 1977),p. 527.                  Arctic Passage-The Turbulent    History of the     by the Aleutin the same manner otherfish, or
                                                                              as
14. Identifiedin earlierinvestigations S. droeba-
                    as          Land and People of the Bering Sea, 1697-1975        (ii) halibutwere not availableor sought during
  chiensis but recently changed to S. poly-        (Scribner,New York, 1975)].                this era. More extensive dissection of the mid-
  acanthusby D. L. Pawson,Smithsonian   Institu-  33. Sea otters reinhabited Attuabout1965.The pop-       den stratigraphy   and more thoroughexamina-
  tion.                           ulation now appearsto be well establishedon        tion of the fish faunal remainsperhaps would
   410                                                                     SCIENCE, VOL. 200
  serve to determine which explanation is true.   47. D. P. Domning,Syst. Zool. 25, 352 (1976).      (1965); J. A. Estes, unpublisheddata. Popu-
40. Thereare presentlyeightspeciesof rockfish(Se-   48. Nearshorecommunitystructureat Adak Island      lation estimates of seals are uncertainbecause
  bastes spp. and Sebastolobus spp.) which have     in the Andreanof Islandsis similar thatat Am-
                                               to        seals are readily observableonly when hauled
  been reportedfrom the AleutianIslands, only      chitka Island in many respects (12, 13). Sea ot-   out, and their haulingout behavioris poorly un-
  two of which (Sebastes ciliatus Tilesius and S.    ters are near carryingcapacityat Adak (K. B.     derstood.
  polyspinus) are abundant nearshore. Sebastes     Schneider,personalcommunication).        53. We thank R. Desautels for access to unpub-
  ciliatus, by its prevalenceand higherabundance 49. J. A. Estes, in Environment of Amchitka Is-       lished data and specimen materialfrom 49 Rat
  in today's communities  (18),is probably spe-
                       the     land, Alaska, M. L. Merrittand R. G. Fuller,     31; E. J. Dixon andthe Universityof Alaskamu-
  cies occurringin the middenremains.          Eds. (TID-26712,Energy Researchand Devel-      seum for providingadditionalmaterialfrom 49
41. K. W. Kenyon, J. Mammal. 46, 103 (1965); T.      opment Administration, Oak Ridge, Tenn.,       Rat 31; and R. Burgner,P. Dayton, D. Eggers,
  H. Scheffer and C. C. Sperry, ibid. 12, 214      1977),p. 511.                    C. Fowler, C. Harris, P. Martin,R. Nakatani,
  (1931); F. Wilke, J. Wildl. Manage. 21, 241    50. The black oyster catcher (Haematopus bach-      R. Paine, J. Palmisano,and C. E. Ray for criti-
  (1957).                        mani), which also preys on limpets, is common    cizing earlierdraftsof the manuscript. McMa-
                                                                          J.
42. P. K. Dayton, in Proceedings of the Colloquium    in the Rat Islands but absentfrom the Near Is-    hon and S. Nancy Steinfortassistedwithlabora-
  on Conservation  Problemsin Antarctica, B. C.    lands.The presenceof oystercatchersin the Rat    tory analysis. The Aleutian Islands National
  Parker,Ed. (Allen, Lawrence,Kan., 1972).       Islandscomplementsthe effect of sea otter pre-    WildlifeRefuge, and particularly R.V. Aleu-
                                                                        the
43. R. I. Black, Quat. Res. (N.Y.) 4, 264 (1974).     dationon limpetsto some unknownextent.        tian Tern,providedessentiallogistic supporton
44. C. A. Repenning, J. Res. U.S. Geol. Surv. 4,   51. C. A. Simenstad,unpublished  data.         Attuduringthe 1976fieldseason. This workwas
  305 (1976).                    52. K. W. Kenyon and J. G. King, "Aerial survey     supportedas a researchprojectof the National
45. L. M. Gard,Jr., G. E. Lewis, F. C. Whitmore,     of sea otters, other marinemammals birds,
                                                and      Fish and WildlifeLaboratoryand by U.S. Fish
  Jr., Geol. Soc. Am. Bull. 83, 867 (1972); F. C.    AlaskaPeninsulaand AleutianIslands, 19 April     and WildlifeService, contract 14-16-0008-2043,
  Whitmore,  Jr., and L. M. Gard,Jr., Geol. Surv.   to 9 May 1965," Bureauof SportFisheries and     to the University of Washington.Contribution
  Prof. Pap. 1036 (1977).                WildlifeReport, on file at the Fish and Wildlife   No. 482, College of Fisheries, University of
46. L. Stejneger,Am. Nat. 21, 12 (1887).         Service, Department of Interior, Anchorage      Washington,Seattle 98195.




NEWS AND COMMENT                                             phasis. Schally seems content to con-
                                                     cede a draw, having written that the
                                                     credit for solving the TRF project "had
                                                     to be shared with Burgus and Guillemin,
          Guillemin and Schally: The                              who elucidated the structure of ovine
                                                     TRH* about the same time." Folkers, on

         Three-Lap Race to Stockholm                               the other hand, says flatly that "We were
                                                     working totally independently of Guille-
                                                     min and his team and we got it before
                                                     they did."
  The discovery made by Guillemin's        three amino acids were evidently not          The TRF molecule did not respond to
team on the eve of the January 1969 con-      joined together in any simple way or
ference in Tucson was a small step for-       Schally would have solved the structure
ward in one sense, a major advance in        with one of his synthetic tripeptides in          This is the second of three arti-
another. After processing some 270,000        1966. If the new composition were an-          cles on the history of the pursuit of
sheep hypothalami they had obtained a        nounced at the Tucson conference, the          the brain's hormones by Roger
1-milligram sample of thyrotropin-releas-      prize of deciphering the structure would         Guillemin and Andrew Schally.
ing factor (TRF), the hormone with         be up for grabs by any chemist in the          Last week's article described how
which the brain directs the pituitary's       world, with the Guillemin team having          the two scientists had spent 7fruit-
control of the thyroid gland. Their         only a 3-week start.                   less years in search of the putative
sample was pure enough to allow two                                     hormone known as CRF and a fur-
                               A Photo Finish Race for TRF
conclusions to be drawn. First, the sheep                                  ther 6 years in quest of TRF. To
TRF molecule consisted of three amino         Guillemin took the gamble and an-           decide whether to continue sup-
acids, glutamate, histidine, and proline-      nounced the composition. In the event,         porting research in the field, the
the same trio that Schally had found in       his start was more than abolished.           National Institutes of Health con-
1966 in his preparation of pig TRF.         Schally, who had temporarily abandoned         vened a conference in Tucson, Ari-
  Schally had had chemists at the phar-       the TRF problem, instantly perceived          zona, in January 1969. Three
maceutical house of Merck Sharp &          how close his rival was to the coup of         weeks before the conference be-
Dohme synthesize the six possible com-       being first with a chemical structure for a       gan, averting an otherwise almost
binations in which the three amino acids       brain hormone. At the conference site he        certain cutoff of funds, Guillemin
could be arranged. (He declined to share      joined forces with an eminent structural         obtained a result of critical signifi-
the samples with Guillemin on the          chemist, Karl Folkers of the University         cance.
grounds, says Guillemin, that "the FDA       of Texas at Austin, and arranged for the
did not allow such transfers across state      synthetic tripeptides to be transferred-
lines.") But all six tripeptides were bio-     across several state boundaries-to          the established chemical tests for identi-
logically inert. Schally had therefore       Folkers' laboratory. Guillemin, also in a      fying the ends of peptides, so evidently
concluded that the biologically active       call made from the conference, asked         nature had blocked the ends in some
part of the hormone must reside in the       Hoffman-La Roche to synthesize the six        *The two teams naturallyhave differentnomencla-
other two thirds of the molecule, with       tripeptides which Schally would not         tures for the hypothalamichormones or factors.
which he could make no headway.           share.                        Schally now calls them hormones, which indeed
                                                     they are; Guillemin prefersthe termfactor to distin-
  The second conclusion which Guille-         From January through the fall of 1969       guish them from all the other hormones.The term
min was able to draw was that the other                                 factor was first used by Saffranand Schally in the
                          there ensued a furious race to solve the       name CRF. There are differentversions as to who
two thirds didn't exist-it was just an im-     structure of TRF. The finish was so close      coined the term. In their respectivecontributionsto
                                                     Pioneers in Endocrinology, vol. 2, a forthcoming
purity, the three amino acids being es-       and confused that to this day both teams       volumeof memoirseditedby JosephMeites, Saffran
sentially the whole of the molecule.        claim priority, although on the Schally       says that "we" coined the word CRF, Schally that
                                                     "I" did, and Guilleminthat creditbelongs to R. A.
  But now came a hard decision. The        side with some internal difference of em-      Cleghorn,anothermemberof their department.
SCIENCE, VOL. 200, 28 APRIL 1978                              Copyright? 1978AAAS
                                0036-8075/78/0428-0411$01.00/0                           411