Progress Report Series.No. 39April. 
l948
A G R I C ULT U R A L -EXPEIMENTSTATION
or The Al abamrn a Polytechnic 
uI ne urn, Aa.
M M. J. FunchessDiredor DI.
.SXZE and SHAPE of 
PLOTS -ad D"'T UO 
f
YIELD for FIEL E7X 8:'With PE'ANUT
T. A, BancroftL/, Coyt-WIsor 
and /
toL '.L ar -."h to:.Co iuct -.(q- utur
Experiments-are often done in ord'r to lean
experimnent, niforinity trials are eampc 
Cinpnt
merits are reported. in the literaturre of','-gricultUr d A00ataogue giing
the location of the da, ta accoaynga 
uch trils aplei1
The only avail able info-0 rmation .to.date co i th 
imd, sze
an shape of plots for rpeanut vwariety tost inwab publshdb attie
et al (2),w TIheir prk- ii rowas madnly conceed .with 
pan
conclusion was tha a plot qf100. square feet, ''consIng Lof ig
100foot row or five or si 2O-foot rows, 3 feet apr (a satis1actory.
Thedaa.prs~r~bd--he refrm In xprion cnducted: solely 
to
eerie optimum size and shapof 'plOt for futureLpeanutexperi-
en-t~aio, n addition to furnishinJg data for t St o z
sha-lpe of. plots, uni fomityr da-lt a also f urnishe a sple 
e.fr equenc c istribu
tion that _ma-y be tested for deplirture from nj rmialiy
No -attemapt was 'Imtad e t o va@ry":th enim ber..of tr eatents' 
or the shapes
of thle replicates, ince 
the, axounit o dta wasoflit 
ed'.extn Ol
the randoized block design- was assumed. oattmptwa s 1i lc. e to 
determine
the rela-,tive, efficiency onl the basis, of -,cost's'Las w-vell 
as land,-use..,
MIATERIAL and 11.MTODS"
H areawas plntedtocon te renmeand rceved40
iTDirector, Statistical Laboramtory, Aaam 
oltcniLnsiue.
SPlant Pathologist, Alalbama Ari-cultural Experiment 
t tion.
~/SuerntndntSSirgrsubstat-ion*,-Alabm gricul 
tural .Exper iiiient
StaOtion.
heThe nmbers in parentheses refer to litraur 
ted
of )100 poun(1s per acre of 0-12-20. The peanuts wer --ante -1 t _ orse-
dravm planter. , hle first twvo cultivations were itI- a horse-drawn rotar
oe; afterwrs the t7Vro parts were cultivated wKith a mule-drvmcultivator,
The area, was considered to have received an averaemount ofr f luring
t he growinm season The peanuts were pickod byr hand on September 12,1046
and green weights in pouncs were recordedv- b 1y plots. achplot cons'sted of
16-2/3 linear feet of rowi and was 3 feet in vie th, making a plot areaof
50 square feet, Part I and Part 
II each centained 108 plots.T
tri al (ffers from other uniformitiy trials veperted in the literature in
Lzit it arranged as a eun-ica -it;e uniifomiity trial, i.e., any result obtained
f rom Part I may be chicked for the particular ye.r under cons*ideration again-
''st the result froli Part II andvice.Versa, The txro parts were taken as
repros entative sape of tie epeiinta1 ar.eaf a s ndtopeanutsa
In order to s t udy th-le distribution of wcightsfreruency tables were
maf10,de for both Part I and Part II; anl thfe st.tisti cs g_ and g, measuring
skewness and kurtosis were caplculated for each (3) The standard de-
vi-ations of g, and *9 were also calculated i thesestests
of significance can be made to measure the denarture from norml
the two distributions of weights
One assumption in the use of the analysis of'variance for analyzing
iodern statistical designs, is that the popltiUond
The: probabilities associa ited. ui-th the variance ratio is onlysit dis-
turbed for moderate llepartures froi~i normality.
ToC11et erini whe41h ether t h e soilet equally variable In two
dire-ctios, th e rnean. '-cllares bet-vieen rows mayv be compared with the mean
squares between column _f(orbt1:3 ot h Part I and Part 11.
To study- the ue-stion of the miiost efficient shape andsize of plots,
a ranomzed block design as assumed; each block or replication was
assumed -to cointain 6 varieties or treatments(3) T ThidSparticular design
and numnber- of va-,rietie r rAmet er hsen., since certain future
trials -e to be so desig7ned., As- far as-po s sibl e) compct blocks or
blocks approaching a square w,,ere used, Plots w -it'hin a block extended the
whole lIengrthl of the respective block, 1iff'erent shapes anmd sizes of plots
withinrpicton were form-.ed by coiabinat ions of the ui* t plots. IAn
analysis of variance was obtaidned in each case. S-ince here were actually
no differenrces between variedlties or trea, )tments. th-e varinty or treatment
source of 'varia.)tion w,,as -not rem-,oved in the anialyses of variances, Presented
below as a..Ll example., is the analysis of variance f or the design Jiwhich
e a ch1Jplot contains only one unit plot f or- Part 1. In such a case there
will be 18' replica--tions,
tj ct ( --, 'Is a c sua
tabl"o) Vr nes, a corrc.spOndlng table -ra caicula' eof coeffi-cients
of variation or stanMda -rd 2evi1tlolls na m3 e. p 1 bSi S; as)percent-
ages cof thj'e siearns Tpe 1 rocLative effici_'encies iorthe di ffe rent. 5yjpQS.
and.. sizes of Plots were derived 10v dCLLv iJ n { 
1 
Osc-1eur ft ecofcirc
o f V ai3.a-L i cn f or u-he u-Irirbsi.zoc-! 1ot itiUiD OinlIj:nlrrt Of h rt~
o f tinae xdre,,;olJ-Y e 1la rge siz e plo o -b e er ea0oL,?h e un Jssi-.ze Qlotand
thle square of ti c co eIff *icient of varIato o h re r s-'Lz e .aIot ()
In c a ses where ull.11the 1jel d sare used, t he serel-at iv ee f fi:Lc-Len~.cie S aY.
al, o be. obtain-.ed by (lvldIlr4; in turn the variance. for -thIie iu nit'.s iz e .p..ot
by the qutetmfteote aiances and th-)e ntL .abe r of unit s :?e plo0t s
contaidned, in thAe 1laIrger plot', It is easy to se t ht th1.e se tontta r
identAcajl. onsidev the unit -plot design as cuiipar:ed -to the dc!esign l~4itb
plots tw,,ice ti-Ie unit size. Let n be t he t o tazInunbi-er -of plot's 3.i the..
unit plot desi"gn and V and VTC be the wihn okvracesfo the
unit plIot and.thec larj.-er plot ide sig n, rspectivIby 
e 
1  
n be
the respectuive coeffic-1in1.ts of vraion zand S b")e th1-ie 6ov er all 5L
Then
( ) n
2 
V
1
S.
2 2
ord (C) 
n V
2
I-Letve or 2ci(cy nV 23
-[I'n XV
hut, te lrightuside ~eLducne omto w
The weigh'ts f eantsfrm icohl67'n-oo potaV gveSi
Tabl 4'~A ad B lie men niglt ~e 
1
)ot orSPrt i-- V6 an fo
Part I is1 
V2
4
For Prart I. the mean squai e between columns with. 5:degrees of free*-
doiin is a, .13728P while the mreani square between rows writh 17 degrees
22
sieeo geneityl ." in1the0tw ic tinos i gr fP at differ.33204twan
22
F~~~~~~~~~~Ior Part II 
1 
.04adtesadrderro 
2
~S~235
henc ete.0006nitiheint'detinerofderes3f2redo. heefre
To2 72testihin'cte I kutss 
0 
.05, s
2
i.45lar hencet.006 hc nia
For PartI, 
1
)=.0131q, igd2.325tandatrd .00,owhilg wa .059,-" *3
1.ence ,00andwth.0109,iain indictin f lgrn epartureefrom normeloity
aheskaris skeis ntsindiufiarey conernedo.thto henrpi
variaetbletss are9gven5inTable . 11. hc'i c ,,
For'Par IIg, .01 (,A)jVariance b .,06swil/9 for5Part,
Lengths in units of _16-/3
Width in xrows ..2 3 6.
1 *12Q927 o.271107 / 159 115 5147
2 '-3555 15> l.2763.& 2 l
94
l60J- _I
3 ,1l.3,60. 1. 65 'no5~6
6 II .241253
(B) Varances /for PrtI
Lengths in units o 623
TI T 4 4- L,
The strndard deviations, on a single plot basis, as a percentage of
the mean, i.e., the coefficients of variation, are given in Table 2
TABLE 2
(A) Coefficients of Variationi fqr Part 
I
Lengths in units of 16-2/31
Width in rows 1 2... 3 6
1 149651 "105, 9.71 
7,28
2 15.3/ 12.33 10.19
3 12.25 945 7e66
6 755
( ) Coefficients of Variation' for Part II
Legths in units of 16-2/3
Width in rows 1 2
1 16,88 12,61 1157 9.46
2 20.21 16.94 164
3 17,01 14.68 1385
6 919
y Onp a per-plot basis
Result based on 72 unit plots instead of all 108 due to size and shape
of plot.
It will be noted that as the sizes of the plots increase these co-
efficients of variation tend on the whole to decrease, which is a result
noticed on previous uniformity trials with other crops (5). The relative
efficiencies 
are given in Table 
3,
TABLE 3
(A) Relative Efficiencies for Part I
Leths in units of 6LL 2
Width inrows _ _ _ __ _ 236
100% 95o9 75c9 
67,5
2 454l/ 35334.4V
3 47,6 40,0 40a7
6 
70,5
(B) Relative Efficiencies for Part II
Lenaths in units of 16-2/3t
Width in rows 1 23
1 10/ O8 99 6 71 0300
?v32 24-W 17. W
3 32o8 22,0 1695
6n3 _ns _
Resut based on 72 unit plots instead of all 108 due to size and.shape
of plot.
P-i-ecis~ ~egvi1-S ~ ~~3
TABIS 4
(A)z'Weight -of-Green Peanuts in Pounds'per .Unit Plot
'11-,) ,1 row Prt I _
i~o N, ~16;-3"Foot 
SeCtions
RwNA-B c EF
13"
18
2'
212
2"5
.22.
88 22
2043
2 
52 5)0
:2 572 w092.
~44 25
2,43 2,522.,40
2 .2~ ,3
2 C'2.40. 3 0
2,35 2.0
() Weight 'C,: Green Peanuts in Punds per Unit Plot
1
6
.2/3~y1 row, Part II
8t'ections ____
1
12.
142
65 2-21K
98
Al
'9
3 91
26
26
3 15
204
66
204
192
0
1 8
-0
J 90
--lo 8
o.-,
2 2012,12,7
2,,61 2 o4 0 
?
2-5 2,6 17
2t1.0 81 2 2
225 L1,40 r 15'
J.9 1.42 16
v 67 1.58 ~5 6
1,46p " .: 1,~
1,50 1..62 0
Row No',A _
ILt is pointed out that in no case is the relatIveefficiency greater
tjr- tht o thC te unit size pLot, However, icr the plot of1 row ,Tidth amd
2 unid- t ].ength hIn Part I, tlhrere is little loss of'officinyie.the
,.elativC eff icioncy- is about 9-6 per.-- cent, T1tjinlet in Part II vs'
about 0 per cent for the r elaEtIve eefficiency,
Ftom a stuy of the reP tve eff 'i-e ncies, iincreasing
the sidze of the plot along the row 'is mjuch :.oroeffic-int than across the
row. This wuld indicate thaat i arrowr rectan-ular s p en plots are ore
effient T iowever, incr-as'i.ng (t'e lngtht ofi othe plot moreha t;o0ties
the unF-it leng, t1with one now o width, m11ayQ roduce -iC offic oncy a
as 2h nor cent or ore :Lyn tl ow' te smallr i ptat
ef f icient in land use, cther co nsidrations mayr nnd one to select a so-e-
what larger plot. Lower yield i n a plot due to one twom i l
is .los se ri us )in the larger .plot
0  
nissing plants hvc.beenobs-erved-to
occur uuiost frequently at tile end of a plot ro, ie adja t b
space etween blocks or re licatiolis. Tis efect
ported uni._0,formiLLty. trial, since no buffer space was providea betwUoeen -lcks
o ro Cti r;
SUIIIARY
I A nul:niorrdi tytrial. for poanut~s is reported i.
ence w"as noted in soil heergo i in two Circ ;iens.. No epature from
oral y wras noted in tee qrOuCncy diy-strreution Table of relativ
eflf.cIcie-rines indicate th-iat ,a plot iT -the shape of'along narrow rectangle
along te r~o, is more ef fi.c i eni;t plots of ne row d
i.n en itu are te rost eficient, allt h ough plots of the widti but
s o te-Jr iatonrm beP.rce e j.-ed because of such considerations as un
even stanid s I f plo t s -ar e 1in cri-e asc-.e d Inr.leng rth1 to over 33-l/3)f eet wi'th
one ro~w -vr:,_J0th orx more I, 1 then th1-ere m,.,ay be a. m~ter-ial decrease in efficiecy
LITERATURE CITED
(I) CCTJIL,, 111pG')Ct a'lo-ue of Unif ormlity-x Trial Data, Supplement, 
Jour.
-oa tat, hoc,, Vol, lV -No.2 
13
(2 LATIC .J-1OSLV U, N :A E. HT,, Plot and Plant
Variation in Virg,inia Peaniut Ame. Soc. Hort,, Sc'
0
34 586-589, 0936
3)STIL,.'DE-'CORf,. G, 111.Statis,:tiJca]l 1liethocis, TIIowa Sta.IteCleePes
T,th. edition, 19146
(4) i1laif, F .11', Size and-1 Shape of Plot in- n& a'.., ,,tijon- to Field Experiments
ith Sugar ets or gr e ,44 696~132
(3)JUSUSE,3S. I. nfunco)Szrad 
hpe of Plot s on the Preciio
of F E'N IeTld DceofnSith Potatoes-
0  
JurCAr
0
Dc, 2 o66i2
- 712 _1_ C* 0