p01.u:m st1.j!jies at - asu digital repository · work was begun through the chemical processing of...
TRANSCRIPT
o
?
P01."u:m ST1.J!JIES AT
,Jaffl(;l;�:\ SC.h\)i�n· .. ,:nt t.e "
Palyuolot,S Labol'atoY.'Y C)C'!KU'I.:.:lGnt loX Anthl.'opolo,.,(y Ar.i.%,o�l.a State Uni-'!':orsity January 1973
In October of 1972 a �er1es of 60 sediment samples were sU;;>lfli tted
to the Palynology Laborato-t·y of the DepaX'-tmellt of Anthropology at
Arizona State University for archaeological pollen analysis. The
contracting agency, the Arizona Al'chaEwlogica.l Center of the National
Park Sel'vice, .i.'cques ted the accomplishmen t of such analy sas as was
PQssible in the t,:qui valon L of one mon th t s research effort 0 No priorities
were placed. on indi vi. dual samples. Howe vex' , 1 twas reques tecttha t
ulaxlmal effort be d.ir!i-�cted toward the test of a specific hypothesis and
that a specific format of analysis be utili:z:ed insofar as this was
possible.
Wide Ueed Ruin is characterized by the Qccurrence of rooms of
distinctive architectural char<leter. l;)laza areas, Ii kiva, large rooms
and smal l rOOlllli) are among the arch! tectural patterns which have been
excavated. Small rooms, which (!onstitute the majority of the architec
tural ullits, have various types of floor features: some have no floOl'
features a.t all, some have bins in the fl oors . others have flagstone
flooring surfaces. 'l'he hypothesis given priority fo_r testing through
pollen analysis WLts that variations in architectural patterns observed
at Wide Reed Ruin are reflectlons of functional 01: use distinctions of
the rooms, and. that these dist.inctions may be recognized and identified
by the variances they cause in pollen rain50
J�vidence that the hyp01;hesis is testable has been provided by the
work of Hill anci Hevly (1968) at Broken K Pueblo.. At that Site,
palynological research allowed recognition of a close correlation of
pollen rain patterns ant:! architectural pattel'ns. Int;erpretation of the
-2-
pollen rains appeared to support the usage interpret.&. tions and functional
interpretations of B!'oken K architectural patterns which were made on
artifactual groundso However, as has been pointed out by Schoenwetter
(1970:39), these findings were serendipitous.. No research program had
been specifically designed to establish the correlation between pollen
rain patterns of specific sorts and ar(!hltectural patterns, nor was the
palynological study designed to allow statistlca,lly acceptable tests of
the correlation once it was established.
There exist two methods of approach applicable to the problem of
establishing a correlati,on between pollen rain and architectural pattern
which may be interpreted in terms of usage and function. The direct
approach is to isolate a small series of samples representing the pollen
rain of architectural features which are so distinctive that they are
very unlikely to represent members of the same functional or usage
category. For example, small rooms with floor bins as distinct from
k1vas. Samples of floor contac t sediment from rooms of such distinctive
a,rch! tecture would be observed to isolate the pollen rain distinctions
which exist t or to demonstrate statistically 'that no correlations
between architecture and po,Hen rains occur. If the palynological
distinctions were subtle, involving significant differences of less
than 5,,0 per cent variat;l.on. (which is not unlikely), observation of a
sufficiently large nlwber of pollen grains from each sample could be
very time consuming o Since labor costs are high in pollen analYSiS, the
direct approach to resolution ot the problem could become very
expensive. In allY case, few samples would be observed.
The indirect approach is to obtain pollen spectra for a series of
-3-
specimens representative of eoten�1a!!l functionally distinct pategories
of rooms. A large number of pollen grains are not observed in each
sample, but a large number of sam.ples are processed and analyzed. If
correlations occur between pattel.'ns of pollen rains and patterns of
architecture, an assessment can be made of the pro.bability that the
correlations are representative of functional distinctions among room
types. '!'his is particularly true if the. time variable is held essentially
constant. However, if no correlations between architectural and
palynological patterns occur, the pollen analyst cannot conclude that
no correlations exist. He can only pOint out that they do not exist in
the data available to himo 'Ibs indirect approach risks inconclusive
results in order to obtain data from a greater number of samples.
It was felt Judicious to use the indirect approach in the present
instance. This was considered advisable because: (A) the amount of
research that could be accomplished was limited by budget imperatives.
If the lixed amount of funds available had been expended before the
direct approach yielded sufficient data for resolution of the problem
(that is, if a sufficiently large number of pollen grains could not be
observed in the roughly one hundred hours available for research),
nothing effective would have been accomplished at all� If the same
amount were spent on the indirect approach there was more risk that the
problem of room function would be unresolved, but a greater prospect
that the data accumulated would be pertinent to the solution of other
problems.. (8) The indirect approach happened to be the one employed by
Hill and Hevly in collecting data on Broken Ko Utilization of the same
approach would provide an essentially comparable format of pollen data.
-4-
(C) It was hoped that the availability of 1r10l'e pollen records from the
s1 te would pI'ove useful in assessing such other problems as intra-site
chronology.
After discussion of the advantages and disadvantages of the two
approaches with Mr. Mount, and after discussion of alt'ernative
hypotheses that could be tested through pollen analYSiS, Mr. Mount
isolated the Wide Reed samples he had collected which he considered
pertinent to the problem.. Work was begun through the chemical processing
of 40 sediment samples to extrac t such pollen as might be con.tained.
Normal extraction procedures for archaeological and terrestrial
sediments (Schoenwetter. 1962:207-9; Mehringer, 1967:136-7) utilize
75-150 cc volume of sample. Because the samples which were submitted
were normally of less than 100 cc volume, all of the sample was used ,for
the extraction" Thus no opportunity remains for independent analysis of
the sediments by ,:Ulother inves tiga tm:'.
'rime allowed observation of �{3 of the 10 samples, of which 21
provided sufficient pollen for analysis. This (63 per cent) 1s an
unusually high :frequency of "gOOd" archaeological samples; the normal
average is closer to 50 pel' cent productivity (t::.g., Schoenwetter,
1967:93-97), and not infrequently is about 30 per cent (eog., Schoenwet ..
tel', 196'1:29-30). In 19 cases analyses proceeded until 200 pollen
grains--excluslve of tlethnobotallic" types (Bohrer, 1966:2;_ Schoenwetter.
1970:38)--were recorded. In the remaining two cases 148-, and IOO-grain
counts were accepted as adequate. This analytic format provided pollen
frequencies roughly equivalent in sta.tistical valu� to those reported
by Hill and Hevly.
� p: ::e
�
4
6
7
9
20
22
27
'29
30
31
32
"3(A)
:':3 (B)
,1,3
53
55
56
58
59
61
C',J bJ) ::s <II � o
't:l ::s OJ
8;
<Ii CIl o l-< :v
't:l � C C.
�
ii!
6
1
5
14
2
4
1
6
4
15
1
5
1 13
6
12
2
8
5
6
<II -0 l-<
..-I ..c: 4-l
<Jl ::J
CIl r:: .,.; 'M ..... C. ::l
't:l Ql ""0
'M . .::
a. �
CIl ::s � (u 0.
'M r::
.s
27 23 2·1
17
12
26
9 13
6 18
3 14
32 19 18
60 19 28
,14 33 23
47 15
46 21 22
53 33 21
48 27 27
44 9 31
37 13 19
49 15 16
'1J ::l CJ '"' <IJ ::s 0'
2
4
6
7
3
6
2
1
41 16 26 -1
6J 15 19
49 16 27
26 6 21
45 9 23
43 6 19
2
3
3
5
2
III .,..j ", rl <II U
e as I o
& ,.c; u
Q) <Il <lJ
..... CIl o l-<
.0 S
..:
OJ til '"' o
rl '+-; 'M ..; ::s
.0
�
39 43 33
34 17 4
67 35 24
1 23 66 32
24 30 10
22 28 17
37 26 24
46 24 27
36 16 15
46 20 21
22 30 21
47 30 19
75 15 9
40 26 16
55 25 14 ..
·10 27 8
34 37 18
51 10 24
Jl 41 27
39 31:. 30
III 1\1 ;.. o
..... .... ..-f ..... ;::! bO
.... ..:I
<'l 'M <II
.... El <II
.j.J ;..
�
13
5
38
18
6
III «l r::
'M El �
a
5
4
1
(3
5
19 14
13 5
18 6
a >,
'" I
'"
� «l ;..
't:l Q)
-a �
1
1
1
2
4
22 5 16
17 7
10 12
11 6
1
2
1 13 5 6
22
4
12
16
19
12
19
-1
4 3
4 2
6 2
5 1
6 1
6 3
� >,
.;J I .
> (!)
z
. r.q
1
1
2
2
1
III ::l ..., d .0 o (J k al tJ)
1
1
4
1
1
3
<II C.
. >, '"' I
u: '"' W 0.
� s:: o bO >.
.....
&
1
1
1
1
2
1
[Jl W
.0 'M �
1
) .1
3
" Cl.I ()
..... d I .. III d
-a tJ)
.... (J
1
1
<l) CG til o r::
.,..f E! ::s bJ)
,S
1
III .... .j.J s::
is. o '"'
't:l :::
,-/ rl >,
U
1
al U)
N
3
2
1
2
7
6
2
1
2
Q) a o <Ii
..... U
2
2
4
<1
2
1
1
4
5
1
2
2
d .j.J oM .0 " ::s (J :::l
U
o:n z ;;.: o
� z �
,1
5
2
5
N
200
100
200+2
200
148+3
200+4
200+5
200+6
200+9
201
200+6
200+1
2QO
200-1 3
200+-4
200+5
200+1
200+2
200+3
200+2
0 0
ETHNO-
BOTANIC ARBOREAL POLLEN
ARCHITECTURE TYPE
r 5 M A l
lIfith l
flagstone
R flooring
0
[ 0 M S
POLLEN .::-Ct
ROOM ; � SAMPLE ., q,0 Ji 0
NO. .oJ' (J <l I
2 30 l' 2 2 54 :.3 3 , 3 53 4 3 27 i 4 4 33(A) 4 33{B) 4 55 5 II 31 Ii 56 13 58 2 -
13 59 I 2 15 61 2
�\��.
with bins- '3 4 -
L 9 6 A R
9 • R 0 with 7 2 G 0 bins 10 9 -;.. M 1:. S 18 43 2
WEST PLAZA 20 3
81n r o�erior West 'P!O!<J fiG'" 1I 32 6 8,� I"tenor !'lm.8 22 2 2 8elow Floor 29 2 4
Under pct Rm. I
40 !
(,).� .� .;\t. we �t.� :00,0 ':,,).f' Q�e t'-((\ 15% 50/. 0
I.':'l n I
- • • - • • - I -
- I -
-- -
- • - -
- -
- -
- ,
- I
- -
-
- -
- -
• • -
- -
_. -
_. -
- • -
ue
•
�
NON-ARBOREAL POLLEN
Gt. j) .\\Oi \ ...... �\� e� -(,�'9 �\' 0' 'Voet' C,
40 15% 0 20 0 r;-;--; I t I • - -
• -
• J
- • -
- •
• • • • - -
- • -
- -
- • -
- - -
- • -
I •
- -
- - -
• -
- • -
- • -
• - -
- -
G� �� .. �o ," (:)(G «,9'<' O\'t'-t.
40 10"-0 iO% 10% j i.l l.l r.l
- • I
I I I
I I •
I • (
I I
I I •
I I •
I
• I
• •
I
•
I I
I I I
I I
•
•
•
-5-
'fa.ble I l'€<:ords all obsElrvll Uons ml1de, and Figure 1 pre.!ilents t.hese
in the usual "g:;:aphic form of Ii pollen diagram,> The data justify two
conclusions: (A) therE-J is �p'enter variati()l1 in the polle.n record
wi t�.:!.!E. a ell tegory of archl tectural torm than occurs between such
categories� Essentially, the productive sRmples fall into four
arch! tectuX'al ca tegoriHS! small rooms \vith flagstone floori ng , small
rooms with bins t largf3 rooms with bins, and plaza. Ou1y one sample was
analyzed fJ:'om the second and fourth categorleo, however. Thus, there
are actually only two architectural categories fl'om which replicate data
are available. statisLically significant variations were measul'ed at
the 95 per Cf .. mt confidence level by the confiden<.�e interval surrounding
the binomial dlatributiono Significant variations occur in the pollen
l'ocord, typifij�d by the .t't'lcords of � and Cheno-am pollen. Samples
from the floor of the same room (e.g .. , Room 2) may show no significant
d:i.stinction in any polhm eategory; or t.hey may (a.go, Room 4) demonstrate
significant distinctions in more than one pollen category. The general
tendency is for the fo.l."h\or situaUon to be more normal, but since the
latt(H' situation does occur, it is clear that a direct correlation
between architectural form and pollen record cannot occur.
Furt.her, where a number of samples are available for an architec
tural category, variation among the samples IfHlves no room for doubt
r�gardillg a lacl( of correlation. Each room in the "large room with
bins" ca tegory provided a unique pollen record 0
(B) l'hE� sorts of palynological records obtained at Broken K
Pueblo which allowed discrimination of rOOlll function do not occur at
Wide Reed Ruin. At Broken K, architecturally distincti va rooms px'oduced
-6-
records of ethnobotanic pollen types which patterned as to both type and
frequency relative to architecture. At Wide Reed, the ntlmper of
different kinds of ethnobotallic pollen types is not as great as at
Broken K, nor do statistically significant variations in the frequencies
of these types occur. It is tints clear t hat the format of. analysis
used by Hill and Hevly (1968) to indicate the dif,ferential functions of
architectul.'ally distinctive rooms does not accomplish s:i.milar ends at
Wide Reed Ruin. The limited amo\mt of data provided on each sample
(i.e", the 200-gra.in count) is inadequate to the task of providing
clues to a potentially useful format of analysis for indicating the
differential functions of architecturally distinctive rooms.. In the
present casell obviously, application of the indirect in contrast to the
direct approach to analysis was a lOSing gamble.
An at t(::Hnpt was made to sal vage some of the effort invested in the
pollen work by cal culating the arborGal pollen frequency relative to
the adjusted pollen sum of S(!hoenwetteL' (1970:43) 0 '!'his pollen statistic
is designed to allow chronological placement and paleoecological
reconstruction in terms of a generalized Colorado Plateau pollen
chronology. T�ble II shows the results of this excerciseo
The d ata indicate that the floors of all of the rooms sampled aTe
reierentto paleoecological conditions of the same type, and
apparently of the same ageo 'rh� sample collected below the floor of
Room 1 is referent to paleoecological conditions of a distinctive type,
apparently of a pe.riod prior to the general construction at Wide Reed.
However, since this sample contains maize and beeweed pollen it would
apparently not p.re-date occupation of the 51 te 0 The samples collected
Room SaUlplf� Size of Adjusted Arboreal Number Numbt'.lfS Pollen Sum Pollell Frequency (')
2 30+54 399 44.9
3 27+53 396 43.9
<1 33A+:J3B+55 596 42.2
11 :n+56 401 113.8
13 58+59 399 35.3
15 61 200 36.0
9 6+7 299 33.1
10 9 199 23 .. 0
18 43 199 12.5
West Plaza 20 1<18 47.3
0 Floor I
W. Plaza Bin 32 199 51.3 1·"1001' II
Room 8 22 200 50.0
Bin
Under Rooln 1 29 200 21.5 floor
'fable II: Calcula tion of adjusted AI' �.
-7-
as bin fills from the West Pla.za (Floor II) and the conU guous Room 8
contain relatively more arboreal pollen than the room floor samples.
The average AP frequency of these samples (50.7 per cent ) is signifi ..
cantly distinct from the average f'requeuey (39 .. 2 per cent) of the floor
�Hl.mpleS6 '!'his could be a function of the distinctive bin fill
proveniences. I suggest, however. that the data are consistent with an
interpretation of a distinct temporal assignment :for the bin fill
samples. TIlis would be a period prior to the construction of the room
floors, but after th(� time represented by the sample' from the sub-floor
at Room 1.
Assignment of absolute age to these three temporal units is
complicated by a lack of surface sample control pol len records from the
geographic district in whlch Wide Reed Ruin 1s located. If we presume
that the pollen statistic accepted by Schoenwetter (1970:43) as
indicating "mOdern" conditions would characterize surface pollen records
a.t Wide Reed, dating may be attempted by reference to the Colorado
Plateau pollen chrollolqgy. '!'he most reasonable date for the sub-Room 1
samples that might be applied under this presumption is A. D. 1075-1125.
The dates of A. D. 1240-1275 are those which would be most reasonable
for the other samples, with indications that Floor II of the West Plaza
and the bin in Room. 8 date earlier. on this horizono
It .1s recognized that dendrochronological assessments of the beams
used in room construction offer dates following A. D. 1275 for Wide Reed
(Mount, pel's. comm.). '!be presumptions regarding the character of the
control, pollen records, then, may be in error. Alternatively, the dates
for the beginning of sediment accumulation on the room floor surfaces
o
o
'-8-
may indecdi:>e somewhat <:?urlim' than th� dates of beam cutting ;for rool' (
eonstruction, Or the rooms dat�ed by d,"�ndrochronology and /;.hose dated by
pollen may be different l'oom�.
REFERENCES CITED
Bohror, Vo!'sil1u I.. 1966 Pollen Analysis of the Hay f1<>:llow Site l�tlst of Snowflahe,
Arizona.. Interim llescal'ch !t0port, No. 12. Geochronology Labc)rato:dHS, Univ(l.l�sity of Ari2.Q!lU. 'l'l.Icson.
HJ.ll, Jam.t:l!:l Not und H,. H. H0vly 1.968 Pollell ilt 131.' • .)li.en .K hlel>lo� Some N(;;w Int;;;l'p):etation::>o AllIo.d.can
Antiqui".:,y, Vol. ;33, No. 2t PP. 200-210. Salt Lal�t� �ity.
Mehringer, rGtt'l' J., Jr. 1967 Poller. Analysi,,; of th", 1'uh: Sp,':i.ng'13 ::>i te. In Pl�i�tocene
..:ltUi .. ih;l� j,r; SO·,ll;hl'.t'u N(;;vada, 8(litGd by M. \'IJI'mitltJ;ton and D, J�l:U,s, "'Po 130-200. N .. ::vrJda State MuscUin Anthropological �!lp(�rs. No. l:J� (!�:'.l.'�()n City.
--
S'�hoenwette1.', Jnmcs. �.!.):; .. : Pollc:l .:�naly;:;ls of Elj,i;'JtcclI L::..:huc:)log;ical S:i.tes in Arizollu and
}ie,,) l\i,:�'tj,c(). In £!lapi.;c;rs !!! �� I'rehisto.ry of Bastcl'n Arizona. 1. by Pn'11 S. M�n·ti�l, et. al., P'). 168-209:--Ic'I;ldian:� Anthropology, Chi(;al:;� NHtm.U(Trj sto�'y MIlSctlln_ Ch:i.cac�o.
19,31 Pol1(�n Ht�pol't. In:� �� !!..cport: Aml;:rican Bottoms Arc:hacoJ.ogy, cdi t.eu by Melvin L. !�mvl(;):r I P.F. 2D-33.-� Illinois AJ.'c:hneolog1.cal Survey, Urbl1na.
l!:iG'i ;')011e1"l t;urvey of the Chu�jm V;111cy. Iu An Archaeological Survey (.)f the C.ilU";.�,,t Vall(lY and tho Ch::tco Platea.u, New :Jexico, Part I:' ':'\,l!�ul'al 8.:.::i.cur;o �l:udi.esJ by A .. H. Har.cis, Jo Schocnwettcr, [lnd A. H. WaJ'.'l'eni PP .. 72-10::;1.. M\iS€Ull1 2!. � Mexico H.es(:arch Records, No� ,1, S�1nta Fe 4
1970 Archa(�ological Pr.>ll€H titudies on thtj Colorado Plateau. ��� Ant.iqui.!l:. Vol 0 �15, No. 1 (in press).