morphological variation and interspecific hybridization among desmodium intortum ... ·...
TRANSCRIPT
TECHNICAL BULLETIN No. 82 J ANUARY 1971
Morphological Variation andInterspecific Hybridization among
Desmodium intortum, Desmodium sandwicense,and Desmodium uncinatum
P ET ER P. ROTAR and KUAN-HON CHOW
HAWAII AGRICULTURAL EXPERIMENT STATION, UNIVERSITY OF HAWAII
CONTENTSPAGE
INTRODUCTION .
REVIEW OF LITERATURE .
MATERIALS AND METHODS .
RESULTS AND DISCUSSION .
Flowerin g beha vior of the parent plants and their hybrids .Numb er of flowers per raceme .Number of flowers opened per raceme per day .Length of flowering per raceme .Sensitivi ty to day length .
Genetic studies amon g the three species .Stem color .In terno de length .Silver marking on midrib of leaflet .Rugose leaflet .Raceme length .Seed weight .
SUMMARY AND CONCLUSIONS .
LIT ERATURE CITED .
APPENDIX .
Morphologica l observations on the pa rent plant s .Stem varia tion .Leaf variation .Flowers an d flowering habit .Pods and seeds .
3
;-3
4
THE AUTHORS
PETEH P. HOTAH is Associate Professor of Agr onomy, College of TropicalAgr iculture, Univers ity of Hawaii, and Associate Agronomist , Hawaii Agri cultural Experiment Station. KUAN-HON CHOW is a former graduate student,Depa rtment of Agronomy and Soil Scienc e.
Morphological Variation andInterspecific Hybridization among
Desmodium intortum, Desmodium sandwicense,and Desmodium uncinetum'
PETEH P. ROTAH and KUAN-HON CHOW
INTRODUCTION
Desmodium. intortum. (Mill.) Urb. 'Greenleaf' and D. uncinatum. (Jacq.)DC. 'Silvcr lca f' ar c important pasture legum cs in Hawaii , thc wetter coastalar eas of northeastern Australi a, and elsewhere in the tropics. D. sanduiicenseE. Mey. has also been tried with varying degrees of succe ss. D. intortum XD. -sandioicense hybrids hav e been obtained in Hawaii when the two specieshav e been grown side by side.
Th e purposc of thi s study was to investi gate the br eeding behavior andmorphological variation among the three speci es and their F1 and F 1 X F1
hybrids.
REVIEW OF LITERATURE
Hutton and Gray (2) reported the first successful fertile crosses from thecombinat ions D. saruluncense X D. uncinatum and D. intortum X D. uncinatum. McWhirter (3) first reported success ful cro sses between D. intortum. XD. sandioi cense. He reported a high degree of compatibil ity from his cro sses,whereas Hutton and Gray found wid e differences in compatibility amongthei r crosses. Morphological characters observed in the crosses of Hutton andGray appeared to segregate in a normal Mendelian fashion.
Park and Hotar (5) , in detai led studies of D. sandtoicense, found varyingshades of red-co lored Flowers to be dominant to ncar-white-colored flowers;red stern color to be domi nant to gree n stern color; the silver-colored leaflet
1 Part of a thesis submitte d by the junior author to the Graduate Division of the Universityof Hawaii in partial fulfillment of the requi rement s for th e Master of Scie nce degree .
4 H AWAII AGRI CUL T UHAL EXPERIM ENT STATION
marking to be dominant to g reen-colored leaflet ; and that the fl ower colorand stem colo r were linked to eac h other. Th ey also observed cons ide rablevar iation in the color of the flowers and in the redness of the stems.
Hutton (1) and Rota r et a l. (8) have reported on the breeding behavior ofthe three species. Self- and cross-fe rtiliza tion occurred in a ll th ree species .Hutton (1 ) indicated that they would se t seed only after tr ipp ing, whereasHotar et al. indicated that they would set seed without bein g tripped . Th e obse rved va ria tions were pr obahl y due to the enviro nmenta l di fferences betweenthe two locati ons.
McWhirter (4) found a male ster ili ty factor in D. sandioiccnse. Wh en D.sandwicensc was used as the male parent in crosses with D. int ortum, the F,hybrids wer e uniformly male sterile. Progenies from the cross (D. intortuni (j'
X D. sandwicensc s ) (j' X (D . int ori um ) s wer e completely fertile, indicating that D. intortum act ed as a fertil ity restor er.
Th e genus DesTnodiuTn has been shown to comprise of diploid species with2n = 20 or 2n = 22 chromosomes ( 7) . Th e three spec ies report ed her e aredipl oids with 2n = 22 chromosomes.
MATERIALS AND METHODS
Five plants of D. satuluncense, number ed 5 11, 521. s;-n, 541, and 55 1 ; eightplan ts of D. un cinatum, numbered VI2, V22, U;-{2, V42, lJ52, U62, U72, andU82 ; and six pl ant s of D. int ortum, numbered 113, 123, 13:-{ , 143, 153, and163, wer e used for inter specific hybridizat ion in thi s study. Severa l plants werelost during the study due to virus disease. A tabular summary of the mo rp ho .logical cha racteristics of the surviving 15 plants is present ed in Ap pendi xTabl e A.
Stem int ern ode len gth , leaf let width, and leafl et length wer e measured -tothe nearest millimeter ; and raceme length was measured to the nearest 0 .5centimeter. Leaflet pub escence was measured by counting the number of hairsobserv ed und er a micr oscop e within a circle of 3.2 mm diam eter, and thi s inturn was multiplied by 12.5 to ob tain the nu~ber of hairs per em". Stern colorwas grouped int o three categ ories: red, brown, or green. Gro wth habit wascla ssifi ed as upright , int ermedi ate, or spreading. P lant vigor was rated as excellent, good, or poor. Th e lea flet length multiplied by its width was used asa n ind ex of leafl et size, and the rati o of leaflet length to width was obt ain ed
from the leaflet length div ided by the leafl et width.Crosses wer e mad e accordi ng to the method described by Rotar et al. (8).
Fl ower bud s wer e emasc ula ted in the la te a ftern oon the day before they opened.Po llinati ons were ca rried out the next morning. Th e racem es were kept cove redwith fib er pollination bags conta ining wet cotto n to mainta in high rel ativehumid ity.
INTERSPECIFI C HYllRIDIZATION AMONG Desmodium SPECIES 5
Th e following sets of crosses were made:(a) Cr osses among the three species .(b) Crosses among the F1 progenies of the th ree species :
(D . sandwicense c;? X D. int ortum s ) c;?
X (D . sandwicense c;? X D. uncinatum s ) ~
(D . sandwicense c;? X D. uncinatum s ) c;?
X (D. satuluiicense c;? X D. intortum s ) ~
(D . sandioicense c;? X D. intortum e ) c;?
X (D. uncinatum c;? X D. intortum 6 ) 6(c) The 10'1 progenies of (a) and (b) were seHed and advanced to the F2
gene ra tion. Su ccessful crosses wcre difficu lt to obtain. No ba ckcrosses weremade. SI seeds were ob tained by ba ggin g the raceme s of F , plan ts and allow
in g the flow ers to se t see ds .Flowering responses to daylen gth , germination of hybrid seed, per centage
of .and relationsh ip of poll en a bo rtion , and pod for mat ion wer e com paredamong th e parent pla nts and their F ll F 1 X F l , an d F 2 progen ies.
Ob servations on flower in g behavior were made on (i ) number of flower sopened per ra ceme per da y, ( ii) number of da ys per ra ceme to comple teflowering, ( iii) number of flow ers op ened per raceme, and ( iv ) length ofracemes. Percen ta ge of po llen abortion was determined by counting a t least500 po ll en gra ins from eac h of two flowers per plant. Poll en grains were classif ied as eithe r normal (fu ll ) a nd sta ined with acetocarmine or shr iveled andun sta ined .
Cha rac teristics used in ge ne tic studies were stem co lor, in ternode length,lea flet s ize, lcaflet marking on the midrib, ru gose leafl et, ra ceme length, andseed weight. Stem color was seg regated int o two classes: ( i ) red or brown,or ( ii) green . The 5th int ernode fr om the stem tip of 10 different stems wasmeasured to the nearest 0.5 em for int ernode length . The midd le leafl et ofthe 5t h or 6th leaf from each of 10 stems was measured for length and width
to the nearest mi llimeter. Hacerne length was an average of 5 ra cemes perplant. Hacerne measuremen ts were made when the last few terminal fl owerswer e openi ng. To determine seed weight , 100 seeds wer e weighed to the nearestmi lli gram and then multi plied by 10 to obtain 1000-seed wei ght.
RESULTS AND DISCUSSIONFlowering behavior of the parent plants and their hybrids
Tw en ty-seven plants incl udin g 6 parent p lants, 8 two-species hybrids, and 11three-species hybrids were obser ved for flower ing behavior during Februarya nd March 1967. Obse rva tion s on number of flowers per ra cem e, fl oweringperi od in days per ra ceme, and number of flowers op ened per da y arc presented in Table 1.
0\
TA
BL
E1.
Flo
wer
ing
beh
av
ior
of
thre
eD
esm
od
ium
spec
ies
an
dth
eir
F,
an
dF
,X
F,
hy
bri
ds,
Feb
rua
ry-
Mar
ch1
96
7
Par
ent
plan
t,N
o.of
flo
wer
sop
ened
Flo
wer
ing
peri
odN
o.of
flow
ers
open
edF,
orF
,X
F,
per
race
me
per
race
me,
inda
yspe
rra
cem
epe
rda
yhy
brid
no.
Ave
rage
Ran
geA
vera
geR
ange
Ave
rage
Ran
ge::z:
: >D
.sa
ndw
icen
se::E
531
......
......
......
......
......
......
......
......
......
......
.....
50.8
42-7
68.
86
-12
5.7
3.9 -
7.7
~ >D
.un
cina
tum
C"l
U32
......
......
......
......
......
......
......
......
......
......
.....
46.8
28--
8110
.88-
-14
4.3
3.0
-5.8
I~U
62...
......
......
......
......
......
......
......
......
......
......
..46
.034
--63
14.0
9 -1
83.
32.
6-4
.2c ::=
D.
into
rtum
I~12
3...
......
......
......
......
......
......
......
......
......
......
..27
.819
--H
5.2
4--7
5.3
4.0
-6.8
..-: ><14
3...
......
......
......
......
......
......
......
......
......
......
....
43.3
28--
618.
46
-11
5.1
4.4·-
-6.:
;.15
3...
......
......
......
......
......
......
......
......
......
......
..30
.824
--41
7.2
6-8
4.2
3.3-
-6.8
::: I~F,
hyb
rids
12 -l
511
X12
3...
......
......
......
......
......
......
......
......
36.8
28-4
86.
65-
-85.
64.
0-7
.4"J
J-l ;>
531
X12
3...
......
......
......
......
......
......
......
......
58.4
36-7
79.
07 -
11
6.5
4.5
-7.7
j
511
XU
22...
......
......
......
......
......
......
......
....
40.8
28--
5710
.88-
-13
3.8
2.5
-5.2
i
511
XU
62...
....__
______
______
______
___...
._....
......
....
43.8
26-6
28.
06-
115.
53.
1--8
.7~
521
XU
82...
......
.__.._
___...
...___
__....
......
......
....
46.4
28-5
67.
25-
-96.
45.
3-13
.0--
l ,.,53
1X
U22
......
......
......
......
..._..
......
......
......
.54.
633
-75
12.2
7-2
04.
53.
7-6
..1~ ~
551
XU
42...
......
...___
___._.
__..__
._____
._....
......
....
29.8
27-3
36.
05-
-75.
03.
5-5.
6~ :::
1-",
XF
,hy
brid
s:::
(511
X12
3)
X(5
31X
U22
)___
______
___.2
9.0
1 9-3
67.
84-
103.
73.
2-1.
7::: -<
(52
1X
123
)X
(511
XU
22)
___.._
___...
.71.
05
6-1
0125
.82:
3-29
2.7
2.0
-4.:-\
g;(5
21
X12
3)
X(5
11X
U22
)__.
__....
....5
3.0
44-Q
715
.01
1-2
03.
52.
2-5
.0I:: N
(52
1X
153
)X
(53
1X
U22
)__.
......
....5
1.8
31-
6312
.08-
194.
43.
3-6.
3;:. j
(52
1X
123
)X
(521
X0
72
).._
______
_...2
9.0
21-
437.
44
-11
3.9
2.1-
7.0
~(5
21
X12
3)
X(5
21
XU
22)
____..
.__...
.52.
041
-66
15.4
12-1
93.
42.
6-4.
7;:. :::
(52
1X
153
)X
(05
2X
133
).__
__..__
____4
3.6
29-5
914
.68-
15
3.0
1.9-
3.9
~15
31X
123
)X
(U5
2X
133
)__.
.__.__
____
40.0
26-6
615
.09
-19
?~
i.r-s
.r-
.1.....
(55
1X
123
)X
(53
1X
U22
)___
_____.
.__.4
9.0
31-6
012
.010
-15
4.1
2.1-
6.0
~
(551
X12
.3)
X(0
52
X12
3)
______
______
.62.
244
-78
27.2
21-3
22.
31.
9-3.
2~ c
(55
1X
123
)X
(511
XU
22)
____..
.____.
.30.
021
-38
10.0
6-17
2.9
2.Q
-4.7
~ ~ '" t"l n t"l '" -.l
H AW AII AGRICU LTU HAL EXPEHIMENT STATIO N
Numher of flowers per raceme: Amo ng the parent plants, the a verage num
ber o f fl owers per raceme va ried fr om 27.3 to 50.8 with a ra nge of 19 to 81[lowers. D. intortum had fewer fl owers per raceme tha n the other two species,Am on g the 1" , hybrids, the average number of flowers per raceme va ried from
29.3 to 53.4 with a ran g(~ of 26 to' 77 fl owers. Amo ng the th ree-species hybrids( I-', X F , ) , the avera ge number o f flowe rs per raceme va ried fr om 30.0 to
71.0 with a ran ge of 19 to 101.
Numher of flowers opeuerl per raceme per clay: Amo ng the parent pla nts,
the av erage number of flower" opened pl'r da y per raceme var ied from 3.3 to5.7 with a ran ge o f 2.6 to 7.7 flowers. F, hybrids had an average of from 3.8to 6.5 [low ers with a ran ge of 2.5 10 ~L7 . T he three-sp ecies hybrids averaged
2.:1 to 4.4 flow ers with a ran ge o f 1.7 to 7.0. The th ree-species hybrids hadfewer flowers opened per raceme per da y than the F , hybrids or their parent s.
Length of flowering per racem e: Th e par ent plants ave rage d from 5.210 14.0 days pe r ra cem e with a ran ge of (j to 13 days . 1-', hybrids averaged
from 6.6 to 12.2 days with a ran ge of 5 to 20 days. T hr ee-sp ecies hyb rids av('rag ('d 7.4 to 27.2 days with a ran g(' of -1. to ;12 days. Tl m -o-sp ecies hybrids
took mu ch lon ger to com plete flowering than the F , hyb rid s or their parents.
Sensitivity to daylenA"th: In vl'st igal ions of f lowl'f ing sens itivity to dayl ength
of the parent plants and some uf their lh n ·(··"p t·,·ie" h;'hr ids were made regu larl y during the period Se ptember 1966 th ro ugh Jul y 1967. D. saud ioicenseflowered throu ghout the peri od. D. un cinatum. started flowe rin g in Oct ober
and ended in April. D. intortum sta rted flowering in December and ended in
March . D. un cinutu m sta rted fl owering 2 months r-ar lier a nd ended 1 month
lat er than D. iutortu TIl . Hotar 1'1 a I. ( 8) have shown that D. sanduiicense isin rk-terrninat e. Wang ( 9) has shown that D. int ort um is a short -day plant. Fromgeneral obse rva tions on D. uncinatum , it can he co ncluded th at it is also a
short -da y plant. Th e la rgest number of h ybrids were n owerin g during theshort -duy per iod , wlu-n 5:) to 56 of tl)(' 6 '1. plants wer e in fl ower. During the
summe r months, 24 to :1() of the 64 plants wer e in Hower . Eight of the 64 fail ed
to flower. These wer e all dwarfed with shortened int ernodes and few sterns and
leav es. It a ppea rs possihl e to ex tend the flower ing period o f D. uncinatum orD. lnt ortum. in to the summe r season by hybridization and select ion pro cedures,
Crosses and environmental effects on crossing behavior
Seed se t from crosses: Hesult s of crosses am on g the th ree -spec ies ar e pre
sented in Table 2. In makin g two-sp ecies combina tions , 149 pods wer c ob
tained from 162:~ pollinations. Pod formation was 9.1 percent , with an averageof ;1 .7 seeds per pod. Among the parent plants, open-pollina ted pod formation
va ried from 16.9 to 51.1 per cent , with a verage number of seeds per pod ran g
ing from 3.5 to 6.3. Percent a ge of pod form at ion and average number of seeds
]T
AB
LE
2.
Po
dse
t,se
ed
set,
an
dse
edg
crm
inati
on
am
on
gin
ters
pecif
ich
yb
rid
so
fth
ree
Desm
od
ium
spee
ie.•
No.
ofN
o.of
No.
ofN
o.of
See
dfl
ower
spo
dsP
odse
tse
rris
see
dsge
rm
ina
tion
Cro
sspo
llina
ted
form
ed(p
erce
nt)
per
pod
germ
ina
ted
(per
cent
)
D.
sand
wic
ense
XD
.in
tort
um...
......
......
......
......
...50
772
14.2
3.8
190
70
D.
san
dtc
icen
seX
D.
unci
nat
um...
......
......
......
......
.37
257
15.3
4.0
8236
D.
iruo
rtum
XD
.un
cina
tum
......
......
......
......
......
....
306
72.
3'1.
77
37
D.
unci
nat
umX
D.
into
rtu
m...
......
......
......
......
......
.10
25
4.9
3.2
159-
1
D.
into
rtu
mX
D.
sand
ioic
ense
......
......
......
......
......
336
82.
43.
512
.I:~
(D.
sand
toic
ense
gX
D.
into
rtum
C;)
gX
(D.
sand
toic
ense
gX
D.
unci
natu
mC;
)<5
....
1593
875.
62.
1112
751
(D.
sand
wic
ense
gX
D.
unci
natu
mC;
)g
X
(D.
san
dw
icen
seg
XD
.un
cina
tum
C;)
C;...
.22
010
4.6
2.3
835
(D.
san
dw
icen
seg
XD
.in
tort
umC;
)g
X
(D.
unci
natu
mg
XD
.ir
uort
umC;
)C;
......
..93
337
4.0
.3.0
3936
2: .., ~ :ll
U>
":l
t'1 o :1 n ::= --: tIl ::: s N ;:. .., (3 2: > '" (5 " o ? (;, ~ ~ :::: ~ U>
":l~ " t'1 U
> 10
10 H AWAII AGRI CULTU RAL EXPERI MEN T STATION
per pod declin ed conside rably from the two-species hybrids to the three-sp ecieshybrids. In making three-species combi nations, 2746 pollinations were made and134 pods wer e obtai ned, with an ave rage of 2.7 seeds per pod . Percent age ofpod formation of three-species hybrids va ried from 4.0 to 5.6 percent , withaverage nu mber of seeds per pod rangi ng fr om 2.3 to 3.0.
P er cen tage of pod formation from crosses was always lower than from theparenl. Cross compatibility varie d am ong the types of combinati ons mad e. IfD. sandwicense was used as the female parent in crosse s with D. intortum and
D. uncinatum, the per cent age of pod format ion was relatively hi gh , 14.2 per cent and 15.3 percent , res pective ly. If D. sarulioiccnse was used as the maleparent s the percentage of pod for mati on in crosses was low, 2.4 per cent.
P er cen tage of pod form at ion fr om crosses among D. intortum and D. uncinatum was low, from 2.3 to 4.9 per cent. The var iability observed from onecro ss to the next suggested that physi olog ical fact ors wer e influ encin g pe rcentage of crosses.
Environmental effects on crosses: T he rel atio nsh ip hetween pod fo rm ationfrom crosses and monthl y temperatures is presented in Fi g. 1. Higher temp er atu res resulted in poorer pod se t. Corr ela tio n between percentage of pods setfrom crosses and the monthl y aver age Iempe rnlu rcs during the 1965-1966winter season was negat ive and sta tist ically highly signi ficant, r = - 0.997,df = ;1. Over 500 crosses wer e mad e eac h mont h, fr om Decemb er to March,and 200 were made in Apr il. Th c res ults a re in agr eemen t with those of Parkand Hotar (5) and Hotar et a l. (8) who indic ated that the best results wereobt ained duri ng cool weather with high relative hum id ity. Th is is also inagreement wi th Hutton (l) who found that poll en germination of D. uncinatumwas poor when the relative hu midity was low.
Seed germination: P ercen tage ger mination of open -po llina ted seeds fromthe pa rent plants was usua lly higher than 95 per cent (all seeds were scarifie dby pri ckin g the seed coa t with a m-cd lc}. Percent age germinat ion of hy bridswas much lower . In Iwo- and three-species hybrids the pe rcen tage germinationvaried from :)7 to 9 tl percent , and f rom ;15 to 51 per cent , resp ecti vely. Althoughthe three spec ies arc closely relat ed as was indica ted by pod set and seed set,the buildup of incompatibility diff er ences du e to possibl e ch romosoma l differences resulted in a lowerin g of seed viab ility in goi ng from two-sp ecies tothree-sp ecies hybrids.
Results of germina tion tests fr om open-p ollinated seeds of the parent pl antsand their two- and three-species hybrids a re presen ted in Tabl e ;1. Th e resultsare compara ble to those present ed in Appendix Tabl e A. Per centage ger mination decreased consid erably, fa lling fr om ove r 97 percent for parent s, to 56per cent for two-species hybrids, and to 41 per cent for three-sp ecies hybrids.
IN T ERSPE CI FI C II YBHIO IZATl ON AMONG Desmodium SPEC IES 11
7 72
\ u,0
6 \ 70z \ lLJ0 \ cr
::J
~\
~\:E 5 \ 68 crcr \ lLJ0 Q..u, \ ",.. ........ -. :E
\ " ...... -- .... wc \ ./ I-0 4 ./ 66\ /Q.. /
~'---------J /u, :t:0 I-
3 64 zlLJ 0C) :E~Z lLJlLJ 2 % OF POD FORMATION 62 C)
0 e:t0:: - - - - - AVG MONTHLY TEMPERATURE 0::lLJ lLJQ. >
60 e:t
APR1966
MAR1966
FEB1966
JAN1966
OL--__-L. L..- -L. ---I
DEC1965
FIG. 1. Relationship between pod forma tion fr om Desmodium crosses and monthly tempera tures, December 1965 thro ugh April 1966.
12 H AWAII AGRI CULT UHAL EXPEHIMEN T STAT IO N
TABLE 3. Percentage germim't ion of open-pollinated seeds a",o ll l-( th r ee Desmotl iurnsp ec ies a s enrupured with see ds from their F , and F, X F , h "hrids
Pl ant or cross
Pa rent plants
D. sand uiicense
D. uncinatum
D. intortum
1', hybrids
D. satu l uiiccnse X n. intortum
D. .sanduiicen se X D. un cinatum
D. in/or/urn X D. un cinatum
D. uncinatum X D. iT/ IO IIUrn
D. lntortum X D. sanduiicense
Seed germina tion (percen t )
99
96
96
7036
37
94
43
1", X 1', hybrids
(D. sandtoicc nse X D. into rtum) X (D. sanduiicense X D. un cina tum) 51
(D. sutuluiicensc X D. uncinatum} X (D. sandwicense X D. illtortum ) 35
( D. sandwicense X D. intortum} X ( D. uncinatllm X D. int ortum) 36
TABLE 4. Average percen tage pollen ahortion among three Desmodium speciesa nd their F , and F , X F , h yhrids
Pl an t or cross
Parent plants
D. sandwicense
D. uncinatum
D. intortum
1', hybrids
D. sandwicense 'jl X D. un cinatum. sD. sandwicense 'jl X D. intortum ~
1', X 1', hybrids
(D . sandioi cens e 'jl X D. intortum ~ ) 'jl X
(D. sanduiicense 'jl X D. uncinatum ~ ) s(D. sandwicense 'jl X D. un cinatllm ~ ) 'jl X
(D. satul toice nse 'jl X D. intortum ~ ) ~
(D. satul toicense 'jl X D. un cinatum ~ ) 'jl X
(D . uncinatum 'jl X D. intortuni ~ ) e
Average poll en abo rtion(pe rce nt )
2.5
4.5
2.6
2.5
22.1
19.3
1.8
10.9
IN TERS I'E CI FI C 1/ YBHIIJIZA T IO N ,\ MO NG Dcsmodium SI'ECI ES
Pollen ab ortio n : Averag es of pollen ab o rtion stud ies are presented in Table
4. Pa rent pla nIs bad less tha n 7 (WITI-nl a bo rted pollen . D. nncinatum had ahii!her pe rcI ~nl age of pollen abortion tha n D. sundioicens« or D. intort um, I nthe F , genera tion. hybrids o f D. sarul ioiccnsc X D. uncinatum had a low per
centage of a bo rted poll en , ranging from 0.8 10 4 .6 percent. wher eas hyb rid sof D. sarulio icense X D. int or tum had a hi gh per centage of aborted 'pollen,ranging from 2.9 to :~ 6.8 percent . This would indi cate that IIIP rel at ionsh ipbetween D. .sandioiccn se and D. unci nat um is closer than that between D.
satulicicen sc a nd D. intortum or th at ther e is some form of inco mpa tib ili ty occurring. In the th ree-species hybrids, the percentage of a borted pollen va ried
from 0.4 to 62.1 per cent. Fifteen . hybr ids from the combina tio n (D . sanduiicense 'i' X D. intortum J ) 'i' X ( D. un cin atum 'i' X D. in tort utu 6 ) 6averaged 10.9 percent aborted pollen ; the ;-H hyhrids of the combination of ( D.satul ioicensc 'i' X D. intortum 6 ) 'i' X ( D. saudioiccnse 'i' X D. uncinatU III 6 ) <3 a veraged 19.:i percent aborted pollen. Crea ler incompatibilities wer eevident in the lall er cross th an in the form er. Pod set in the th ree D CSIIIOdiulII
species was sta tistically h igh ly significantly and negutivcly cor rela ted with the
percentage of ab ort ed pollen , r = - 0.n 6, df = 2 1.
Genetic studies among the three species
Stem co lor: All plan ts of D . uncinatum had red stems; l) . satul uicrnsc hadred or green stems; and D. int ortum ha d red , h rown, or gree n stems. dependin g
upon the plan t. In de termining genetic ra tios, the stems were seg rr-gated intotwo classes : ( i) red or brown or (ii) gree n. Th ere wer e differences in inl ensity of the red stem colo r. Results of the stem color stud ies are pr esent ed inTabl e 5. Th ese ind icate that stem color in the three spec ies is coniro l led by asing le gl'ne with domi nance for red or brown stems a nd recessiveness forgree n stems. Crosses among red- or brown-stemmed pa rents produ ced red - orbrown-stemmed F\ progeni es. F 2 pr ogenies were also red or brown stemmed.
Crosses am on g green-stemmed parents prod uced green-stemmed V, progenies
and the V2 pro gen ies wer e also green stemmed. Crosses amon g red- or b rownstemmed and gree n-stemmed pa rents pr odu ced only red- or br own-stemmed 1-' ,pl ants. The F2 progenies seg reg ated int o red- or brown-stem med an d greenstemmed plants, with 3 :1 rat ios whose ch i-squa re values had pr obabiliti es from0.25 to 1.00. From th e three-species hybrids, the F , pr ogeni es wer e red or
brown stemmed. The F2 progeni es seg rega ted in to red- or b rown-stemm ed andgree n-stemmed plants, with 3 :1 ratios whose ch i-squa re values had small andnonsignificant probab il ities. Th ese res ults are in agr eement with P ark andRot ar (5) who foun d th at red stem colo r in D. sandio iccnse was dom inant togreen. In in terspecifi c hybrids, McWhirter (3) an d Hutton and Gr ay (2) indio
TA
BL
E5
.S
tem
colo
ro
fth
ep
are
nta
lD
esm
od
ium
spe
cies
,th
eir
F,hy
br-i
ds
an
dF,
pro
gen
ies,
an
dch
i-sq
ua
res
for
go
od
nes
so
ffi
tto
3:1
rati
os
of
red
tog
reen
ste
m
f-' ....
No.
ofF
,pr
ogen
ies
F,
or5,
Obs
erve
dC
alc
ulst
edC
ross
stem
colo
rR
edor
brow
nG
reen
Rcd
orbr
own
Gre
enC
hi-
squa
reP
Cro
sses
amon
gre
dor
brow
n
521
XU
22...
......
......
......
......
......
......
...re
dor
bro
wn
3535
521
XU
82...
......
......
......
......
......
......
...re
dor
bro
wn
3636
Cro
sses
amon
ggr
een
511
X13
3...
......
......
......
......
......
......
...g
reen
3434
.... .... :..5
11X
143
......
......
......
......
......
......
......
gree
n36
36< >-
531
X14
3...
......
......
......
......
......
......
...gr
een
3636
:: >- C")
Cro
sses
ofre
dor
bro
wn
Xgr
een
-5
11X
U22
......
......
......
......
......
......
......
ore
dor
brow
n27
927
90.
001.
00c: r-
521
Xu
n....
........
........
........
........
......
red
orbr
own
2511
279
0.60
7.25
.., §53
1X
U82
......
......
......
......
......
......
......
.re
dor
brow
n12
311
.25
3.75
0.20
7.50
>-
Th
ree-
spec
ies
hybr
ids
t'::
;.< ....(5
51
X12
3)
X(U
52X
133
)...
....
red
orbr
own
2511
279
0.60
0.25
t':"J :::
(52
1X
153
)X
(U52
X13
3)
......
.re
dor
brow
n20
518
.75
6.25
0.34
0.50
:: t'::
(521
XU
82)
X(5
21X
153
)...
....
red
orbr
own
315
279
2.36
0.10
:l ...;
til .., ;;.- .... c3 z
TA
BL
E6
.T
ests
of
ind
ep
en
den
ceb
etw
een
inte
rno
de
len
gth
an
dg
row
thh
ab
ito
fp
lan
tsfr
om
the
cro
sses
(D.
san
dlv
icen
se~
XD
.in
tort
um
~)
~X
(D•
.~andlVicense
~X
D.
un
cin
atu
ms)
il
Inte
rnod
ele
ngt
h(e
m)
Cro
ss1 -
22-
33
-44
-55-
-66
-77
-88
-9
Spr
eadi
ngf
......
......
......
......
......
......
......
.....
32
511
64
12
F...
......
......
......
......
......
......
......
..2.
662.
125.
848.
508.
504.
251.
061.
06
f·F
......
......
......
......
......
......
......
.....
0.34
-0.
12-
0.84
2.50
-2.
50-0
.25
-0
.06
0.94
Inte
rmed
iate
f...
......
......
......
......
......
......
..0
00
110
31
0
F...
......
......
......
......
......
......
..1.
170.
942.
583.
753.
751.
880.
460.
46
f-F
......
......
......
......
......
......
.....
-1.1
7-0
.94
-2.
58-2
.75
6.25
1.12
0.54
-0.
46
Upr
ight
f...
......
......
......
......
......
......
......
....
22
64
01
00
F...
......
......
......
......
......
......
......
....
1.17
0.94
2.58
3.75
3.75
1.88
0.46
0.46
f-F
......
......
......
......
......
......
......
......
.0.
831.
063.
420.
25-3
.75
0.88
-0.4
6-0
.46
Tot
al...
......
......
......
......
......
......
......
......
......
...5
411
1616
82
2
Ch
i-sq
uare
=32
.79,
df=
14,
P<
0.0
1.
Z >-i
t'l ::<l
m ";I t'l 0 ~ r. ::I: -< § N :.- >-i
Tot
al
I~34
;;: 0 Z C'l
15I~ '" ;3 0 I;l.
15I~
. m "C t'l 0 t'l
.m
64
......
(Jl
16 HAW AI I AGIUC ULTU BAL EXP E HI M ENT STA TION
cated that ste m color was contro lle d by a sing le pa ir of genes with dom inan cefor the n·d s lern .
I n ter n o de le n g th: Measur emen ts of internode length s are presen ted in Fi g.2. Th ese varied amon g the pa ren ts as follows : 6.0 to 8.0 em for D . uncinatum,:t 4 to G.:) cm for D. int ortu m, an d 1.1 to 2.9 cm for D. sand uncense. Int ernodeleng ths of the two-spe cies hybrids were intermed iate to those of the pa rents.In the th ree-spe cies hyb rid s, man y gra dations from one ex treme to the othe rof the three parents were obse rved. Resu lts ind ica ted that the genet ic behav iorof internorle len gth of these spec ies was controlled by mult ip le genes .
A test of independence of intern ode len gth with stem growth hab it of 64three-species hybrids from the combination (D . stuu lu.icensc <j> X D. int or tum 6 ) <j> X (D . sandwicense <j> X D. uucin atum 6 ) 6 is presented inTable 6. Results indi cated that these two cha racte ris t ics a re closely rela ted. Itwas conclu ded th at plan ts with sprea d ing a nd int ermed iate growth habits hadsignificantly lon ger intern odes than planl s with an upri ght gro wth hab it.
S il ve r marking on miclrih of leaflet: There are th ree typ es of leafl etma rkings observa ble in 11 11 ' leafl ets of these spec ies. 0ne is a reddish-brownfleck scatte red th rou gh onl thr- ll'afl d s of certain D. intortum plant s. Anotheris the shiny a ppl'a ra n(;(' of lh r: midri b a nd veins of the leafl et, observ ed onlyin D. intortum . This Irait var ies in its express ion fr om nonappearan ce to nearl ycoverin g the k afft. l. Th e th ird is the silver-g ray ma rkin g on the mid rib of theleaflet, whic h is observed in all thr ee species but not in all p lan ts.
All plants of D. satulioice nse an d D. uncinatum used in thi s study had silvermarking on the leaflet midr ib. Fo ur of the D. int ortum plants wer e marked.Hesults of midrib leafl et ma rking stud ies a re pr esent ed in Tabl e 7. In crossesamo ng marked plants, the 1" , progeni es were ma rked and only marked F"progt'n it's wer e produced. In crosses among ma rk ed and nonm arked plants,only ma rked F 1 progen ies were observed, a nd marked and nonmarked Lpro genies were observ ed to seg rega te int o :1 (ma rked) :1 (nonma rked ) ratioswhose ch i-sq ua res had probabilit ies va rying from 0.10 to 0.75 , depending uponthe crosses. Fr om thi s evidence it was concluded that the si lver-gray markingon the midrib of the leafl et was controlled by a sing le pair of genes, withdominan ce for the marked and recessiveness for the nonmarked. Th is was in
agreement with Pa rk and Hotar ( 5) a nd with Hutton and Gray (2) who concluded that thi s leafl et marking was contro lled hy a sing le pair of genes.
Rngose leaflet: Pl ants ha ving th is character istic had distincti vel y shapedleafl ets whic h were also cur ved backward . This cha racteristic was ob
serve d only among the 1"2 pro genies f rom th e combina tion (D . sanduiicense<j> X D. int ortuni 0 ) <j> X (D. uncinatum <j> X D. int ortum 6 ) ~ , where D.
INTERSPECIFI C H YBRIDI ZATIO N AMONG Desm odlun i SPECIES
PARENT
17
4
2
o
6U)..J 4sc 2sQ 0Z
I&.0
dz 12
10
8
6
4
2
0
TWO-SPECIES F,
THREE-SPECIES F,
1.0 2.0 3.0INTERNODE
4.0 5.0 6.0LENGTH IN em
7.0 8.0
Ft G. 2. Internode len gth s of pa rents an d two- a nd th ree-"p,,,-;,-s hyh rids of U. sun dunccn se,D. intortu m, an d D. uncinatum.
TA
BL
E7
.In
heri
lan
ceo
fle
afle
lm
ark
ing
011
the
mid
rib
of
the
leafl
el
of
thr-
ee
Ires
nu
uli
ura
spec
ies
..... ce
F,hy
bri
ds
and
F,
pro
gen
ies
Ch
i-sq
uar
eO
bser
ved
Cal
cula
ted
for
Cro
ssM
ark
edN
onm
arke
dM
ark
edN
onm
ark
ed3
:1ra
tio
P
Mar
ked
Xm
arke
d
511
XU
22...
......
......
......
......
......
.....
......
......
......
..36
036
0
521
XU
22...
......
......
..._..
......
......
......
......
......
......
.35
035
0
521
XU
82...
......
......
......
......
......
......
......
......
......
.36
036
0
531
XU
22...
......
...._.
......
......
......
......
......
......
......
.36
036
0
531
XU
82...
......
......
......
......
......
......
......
......
......
.15
015
0I~ :;::
Mar
ked
Xno
nm
arke
d>
511
X13
3...
......
......
...__.
.__._.
__....
..._..
.._...
......
...._.
.27
725
.58.
50.
360.
50-0
.75
> <:"l
511
X13
3...
......
......
......
......
......
......
......
......
......
.30
627
.09.
01.
330.
10
-0.2
52:
531
X14
3..
_.....
......
......
......
......
......
.._._.
......
......
..29
727
.09.
00.
590.
2 5-0
.50
c::: ti c:::T
hre
e-sp
ecie
shy
brid
see ;:.
-(5
51
X12
3)
X(U
52X
133
)._.
_._...
......
__..._
_34
236
0t'1
(52
1X
152
)X
(U52
X13
3)
......
......
......
....
241
250
~ '":l
(521
XU
82)
X(5
21
X15
2)
......
......
.._..
.__..
360
360
t'1 ee(5
21X
123
)X
(511
XU
22)
._..__
._....
......
_...
200
360
i t'1
(521
X15
3)
X(5
31
XU
22)
......
......
.._...
.___
178
18.7
56.
250.
650.
10-0
.25
:2 .., Vl .., :..- .., (5 :2
INTERSPECIFI C HYBRIUIZATION AMONG Desmodium SPECIES 19
into rtum. occurred twice in the combination. Four out of 19 plants from thi scross had the ru gose leaflet characteristic. This produces a fair fit for a :i(normal):l ( ru gose lea fl ct ) ratio with a ch i-sq ua re of :i.55 the probabilityof whieh fa lls between 0.10 and 0.05.
Th ese four plants flowered during the per iod November 10 April as did D.uncinatum. Their seed se t was good, and their per centage of sh riveled seed waslow, 5.0 to 9.5 percent. The 1000-seed weight s ran ged from 3.73 to :i.76 gra ms,midway between D. uncinatum and D. satuluiicensc, Th eir internodes wer eshort, with len gth varying from :i.1 to 5.5 em. The stems wer e densely cove red
with long, uncinate hairs. Stem color va r ied from gree n to red . Leaflet -sizeindices varied from 12 to 16 . The leaflets, cove red with lon g, dense hairs, vurir-din color from gree n to brown. Th e ru gose cha ra cte r was not observed in any ofthe parent plan ts nor was it observed among any of the F, hybrids from the
three spec ies.
Raceme lengt h: Raceme length is one of the distinctiv e [r -alures of the thr eespecies. As the florets open from base to a pex, the ra cem e don ga l('s and n-achesmaximum len gth when the termina l flow er s o pen. Tlw ra cem e decre ases in siz«as the see ds mature. Average ra cem e len gth among the tbn'l' pari-ut s pec ieswas 24.4, 14 .4, and 12.1 em for l ), uncinutntn , D. sandioic ense, and D. intortum.res pectively. Haceme lengths of F, hybrids am on g the three spec ies wer e as
short as those of the short ra ceme parent . In the F, X 1-' , hybrids. ra ceme
length s exceeded the ran ge of the parent s. Hesu lts oluniucd Wl'r e difficult 10
int erpret and no a ttemp t was made 10 determine the gl' lJet ic behavior ofrac eme len gth.
Haceme len gth was sta tistica lly hi ghl y significantl y cor rr-lau-d with !l'nglhof flowerin g peri od per ra cem e and with to tal number of flow er s per rnccnu-,r = 0.651 , <If = 22, and r = 0.542 , df = 22, n 'SIH,( ·tivl·!y. This i::; l'xpl'c1ed .inasmuch as ra cem e len gth will influence till' number of flowers that an' pr esentper racem e, and the length of the flow erin g peri od per racem e is a fun cti onof the number of flowers per ruccnu-.
Seed weig h t: Th er e wer e disti nct differc llcl's in 1000-s('('d wl'ights amon gthe three species and their hyb rids. Hcsu lts a re pr esented ill Tab le 8. D. uncinatum had the heaviest 1000,sl'cd weighls, 4 .08 gr am s/WOO sceds ; D. satul ioicense had the nex t heaviest, :i .54 gra ms/ l 000 seeds; and D. intortuni the
light est, 1.68 gra ms/ lOOO sel' d~ . Th ese result s a re in ag reeme nt with Hotarand Urata (6) . Differences in seed size wer e obtained amon g the hybrids fr omthe thre e spec ies. None of the hybrid, had 1000-seed weights which exceede dthe mean s of the heav iest or light est paren ts.
Percentage of sh riveled seed from th e F I gene ration of two -species crosses
was not significantly greater than that of the paren ts (t. = 1.80, df = 24,
20 H AW AII AG HICU LTU HA L EX I' E Il IMEN T ST ATION
T ABLE 8. Thousand-seed w(Oi/:hts and p ercenuure of sh ri veled seed from th reeDesmodium species and their F. and F. X F. h)'hrids
P la nt no.
Pa re nt
D. satuluiic ense
1000·seed weight(grams)
Shrive led seed( percent )
53 1 3.54
D. un cinatum
U12 __ . __ :~ .99
U32 _ _.. ......... 4.0I
U42 __ _ _.__ 4.08
D. intortum
133 _ _........................ 1.68
163 1.99
1'\ of two spec ies
2.5
2.5
3.0
2.5
4.5
5.0
5 11
5 11
52 1
53 1
S.1 1
551
531
X U22
X U62
X U82
X U22
X U42
X U42
X 142
2.81
3.46
3.00
a.sr3.35
3.25
2.53
5.0
4.5
:~ . O
6.0
4.5
4.0
,1.0
F, of thr cc species
(511 X 123 ) X (53 1 X U22) a ............._- --_........... 2.97
II ............................ 2.91
c .................--......... 3.06
(5 11 X 123 ) X (54 1 X U22) .............................. 2.54
(511 X 123 ) X (5 11 X U22) ...............__........__... 2.06
(521 X 123) X (5H X U22) .............................. 2.52
(52 1 X 123 ) X (52 1 X U72) .............................. 2.48
(52 1 X 153 ) X (5 .3 1 X U22 ) .............................. 1.74
(55 1 X 123 ) X (5 11 X U62) ....-.-...........-........... 2.45
(521 X U82) X (521 X 153) .............................. 3.28
(52 1 X 153 ) X (052 X 133) ...................__......... 3.73
(521 X 153 ) X (052 X 133 ) ....-_ ............-........... 3.76
15.5
12.0
8.0
14.0
17.5
22.0
16.0
19.0
9.5
9.5
9.5
5.0
P > 0.10). Percentage of shr iveled seed from the three-species hybrids wassig nificantly greater than tha t of the parent (t =10.53, df = 33, P < 0.01).Although the percentage of shriveled seed was not high in the F, generationfrom two -sp ecies hybrids, there was a trend in this direction. Evidently the
INTERSPECI FI C H Yll HIUIZATIO N AMO NG Desmodium SPECI ES 21
three spec ies are sti ll quite closely related, and it is onl y when they ar e combined that larger numb ers of incompatibilities occu r. This same trend wasobserv ed in percentage of poll en ab ortion, Tab le 4.
SUMMARY AND CONCLUSIONSTh ese experiments wer e undertaken to study the breeding and flow er in g be
havior, gene tics, and relationships among the three spec ies : D. saiul ioicensc,D. uncinatum, and D. intortum. D. sandioicense was insensi tive to daylengthand flow er ed a ll yea r. D. uncinatum and D. intortum flo wer ed on ly in theshort-day seas on, from October throu gh April , and from Decemb er throu ghMarch, respe cti vel y.
In controlle d crosses, per centage of pod formation was low, an average of9.2 percent for any two of the thr ee species crosse d togeth er, and an averageof 4.9 percent when F 1 progenies of two species were combined to obtain three
spec ies crosses. \Vhen D. sandtoicense was used as the femal e parent in crosseswith D. uncinatum or D. intortum , per centage of pod formation was relativelyhi gh, 15.3 per cent and 14.2 per cent , respecti vely. Wh en D. satultoicensc wasused as the male- parent, per centage of pod formation was very low, 2.4 per cent.
Germina tion percentages of hybrid seeds were low, 54 per cent for seedsfrom two-species crosses and 45 per cent for F, X F, seeds. Hybrids of D.sanduiicense and D. uncinatum had an average of 2.5 percent aborted poll en ;hybrids of D. uncinatum and D. iiuortum had an average of 22.1 per centaborted po llen .
Stern color of the species was confirmed to be und er the contro l of a sing lepair of genes with red or brown dominant to green.
Midrib leaflet markin g was observed in a ll three species . Th e marking wasconfirmed to be du e to a sing le pair of genes with dominance for the mark edand recessiv eness for the unmarked lea flets.
Rugose leafl et chara cter was fo und in the three-sp ecies hybrids of the combination (D . sanduiicens e lj> X D. intortum &) lj> X (D. un cinatum lj> X D.
intortum & ) & . This appeared to be due to a si ngle pa ir of ~enes with the
ru gose lea flet recessive to the normal lea fl et.Ra ceme length, in ternode len gth of stems, and WOO-seed weight appea red
to be controlled by multiple genes, inasmu ch as tho se of the F , plants were
intermediate to their parent s and those of the F" progenies were near ly thesame as or exceeded the ran ges of the paren ts.
D. intortuni and D. uncinatum had leafl et -size indices (leaflet length X leaf let width ) which were approximately twice as la rge as th ose of D. sandioiccnse.
P er centages of pod s set from crosses wer e negat ivel y correla ted with themean mon th ly temperat u res durin g the short-day season. Th e hi gh er the meanmonthly temperatures the lower the percentages of pods set.
22 H AWA ll AGHI CULT URAL EXPERIMENT STAnON
LITERATURE CITED
1. HUTTON, E. M. 1960. F lowe ring and poll ina tion in Indigo/ era spica tu, Pha seolus lathyroides , Dcsmodium IlI/ cill a/1I111 a nd some othe r trop ica l past ure legumes. Empire .I. Exp .Ag r. 2H :235-243.
2. a nd S. C . GRA Y. 1967. Hyhrid iza t ion betw een th e legu mes Desm odi um in -tortum , D. un cinatum , and D. sandtcice nse . J . Aus tra lia n I nst. Agr, Sc i. 33 :122-123.
3. MeW IIIIITf:II, K. S. 196.'3 . An nual Hep ort 1962-63. C.S .I. R.O. Divis ion of Tropical P as
tures, Bri sbane, Q ueensla nd , Aust ruli a . P . :\2.
4. 1969. Cytoplas mic. ma le ste rility ill Desmodium , Aust ra lia n .I. Ag r, Hes.20:227-241.
5. PARK, SOON J At, a nd P . P . ROTAR. 1968. Ge netic stud ies in S panish clover, Desmodiu msandtoicense E. Mey. I. Inh e ritan ce of flower color, s te m colo r, a nd leafl et mark ings.
Crop Sc i. ~:467-470.
6. ROTAR, P . P ., and U. UIIAT A. 1966. So me ag ronom ic obse rva tious ill Desmodium spedes : see d weigh ts. Hawai i Agr. Exp . Stu. T ech. Prog. Rep. No. 147. 14 p.
7. and 1967. Cytolog ica l stud ies in th e genus Desm odium ; somech romoso me co unts . Amer. J. Bot. 54:1-4.
8. S. J . P AliK, A. Bn oMDEP, a nd U. URATA. 1967. C ro,;sing a nd flo wer ing he-havior in Spani sh c lover , Dcsm odium sandioicense E. 1\1ey., and other Dcsmodiumspecies.. Haw aii Agr. E xp . S ta . T ech. Prog. Re p. No. 164. B p.
9. W ANG, C. C. 1961. G rowth , flowering a nd fo rage produ ct ion of some grasses a nd
legu mes in response to diff eren t photoperiod ". J. Ag r, Assoc. China. N. S. 36 :27- 52.
INTEHSP ECIF IC HYBRInIZATION AMONG Desmodium SPECIES
APPENDIX
Morphological observations on the parent plants
23
Stem, leafl et, flow er , pod , and seed characteri stic s are pre sented III tabularform in App endi x Tabl e A.
Stem variatio n : Stem growth habit varied from upright to dec umbent. Stemcross section varied from ro und, obtuse ang led to acu te angle d. Stem in ternodelen gth va ried fr om 2.9 to 8. 1 cm, depend ing up on the species and the envi ron ment al conditions . Stem color vari ed from ligh t gree n, gre en , brown, and redto dark red, dependin g up on the plan t or the spec ies or hath. Pubescence differ ed grea tly in length, tex ture, density, and sha pe fr om plan t 10 plant ; someplants were a lso glandular.
D. sanduncense: Th e five plants of D. sutuluncense used in this study hadan upri ght habit of growth, and the stems were ro und in cross section. Stemint ernode length va ried from 2.9 to ;1.2 em, the sho rtes t of the three species.
. Stem color va ried from green to red , depending upon the pla nt. D. saud
. wicense had few shor t stem ha irs, which were sometimes glandular.
D. uncinatum: All eight plants of D. uncina tum had decumben t and spread.ing growth habit . The stems wer e round in cross section . T he stem inlerno delength varied from 6.2 to 8.1 em. All of the stems of the D. uncinauini wereeithe r red or br own. D. uncinatum had long, dense , uncinate stem ha irs .
D. int or tum : The stems of the six D. intortum plan ts were mult ifa rio usin morphology. Growth hahit va ried from upright to spread ing, an d crosssection of stems from nearl y ro und to acute angled. In te rnode length variedfr om 3.4 to 6.4 cm. The ste ms' pub escence showed great diversity in amo untand texture. Stem color va ried fro m red 10 green.
Leaf variation: Lea ves of the thrce species wer e tr ifo lia te. Leafh-t s ize variedfrom species to species; mature leaves were larger in summer, smalle r in winter. Th e center leafl et was a lwa ys la rger than the Iwo la teral leafl ets. Leafl etform va rie d from lan ceolate ell iptic to ovat e, depending upon the species orthe plant or hath, with in the species. Lea flet ap ices were either acu minate oracut e and leaflet bases were either truncat e or obtuse.
Leafl et margin s were en tire and often provi ded with a row of ha irs. The si] ver leafl et ma rkin g on the mid ril ; of the leafl et was obser ved in eac h species;although some of the parent plant s of D. intortum wer e unmarked, carryingonly a b row nish fle ck on the leafl ets, and oth ers were entirel y un ma rked inany fashi on. Lea f co lor va ried fr om light green to gree n, and brown. P ubescencediffered conside rably in length , densit y, and sha pe between species and am ong
24 HA WAII AGHI CU LTUHA L EXPEIOMENT STA TION
plants wi th in species . The lower leaf surfaces usua lly were more pubescen tthan the Upp('f su rfaces .
D. sarul ioiccns c: Leaflet -size indices (leaflet len gth X leafl et width) of D.sandioiccns e ran ged from R to 12, an d the leaflet length to width ratios variedfro m 1.74 to 1.89. The D. sandioicense plan ts in this study ha d green leafl ets.Lea flet form was la nceolate ; lea f apices wer e acu te; leafl et bases wer e obtuse,and const ricted abruptly to pet iolules.
D. uncinatum: Leafl et-size indices of D . un cinatum. varied from 21 to 27;the leafl et length to width ra tios varied from 1.77 to 1.86 . All of the D.uncinatum. pl ants used in thi s study had ligh t-gr een leaves with a silvermarking a long the midrib of each leafl et. Leafl et form was elliptic . Thewidth was about one-half the length. Leafl et ap ices were acute ; leaflet baseswer e truncate.
D. intortum: Leaflet-size ind ices var ied from 18 to 32 and leaflet length towidth ra tios varied from 1.49 to 1.69. D. in to rtum. plants va r ied in leaf colorand leafle t marking. 123 had brown lea flets and a silve r leafl et markingalon g the leafl et midrib. The other pla nts were green and lacked the midribmarking. Leafl et form was ovat e, with the broadest pa rt below the midd le.Leaflet api ces were acut e to ac umina te a nd leuflet bases were truncate.
Flowers and flowering hab it: F lower color varied from nea rl y white,th rough pink and red, an d light bluish white, depending up on the species andplant or bot h. Haccme length vari ed among the th ree species. Abo ut 2 weekswere required for flowe rin g to be completed on anyone ra ceme; flower num be r per raceme varied amon g spec ies and among plants within species.
The th ree speci es flow ered in a simila r man ner. The day bef or e the fl owersopened , the closed pet als expa nded and wer e obs erved projecting beyond thesepals. In the la te a ftern oon of the day before the flowers op ened, the an therswere white or very light yellow. By about 8 I'M , the an the rs had turned yellow,ind icating poll en maturat ion. Abo ut midn ight the an thers had dehisced . Bydawn the peta ls had completely expanded but the flo wer was still clo sed. Shortlyafter daylight , the standa rd peta l became erect and the flow er was rec eptiveto pollinating insects. Po l len was usuall y discharged in a cloud a round thestigma when the flower was tr ipped . T he stigma did not s tr ike the standardpe tal as in alfalfa . Once tr ipped , the flow ers qu ick ly wilted, and the standa rdpetal s wer e folded over the stigmas in a short tim e. Two fl owers usuall y occurred a t each no de, and the pa irs wer e a rra nge d in a spira l on the ra ceme .Onl y a few flow er s op elled per r acr-rne per da y.
D. satulioi cens e : Flower color of D. sandioicense varied from nea rly white,
throu gh pin k to red, dl'JH'nding upon the plant. Haccmes were about 15 ern
INTERSPECIFI C H YBRIDIZATION AMONG Desm odium SP ECIES 25
long with about 50 flowers per raceme. Fl ower ing peri od pcr raceme aver aged 8.8 da ys with 5 .5 fl ower s opened per raccme pcr da y.
D . un cinatum: Fl ower color of the six D. uncin atum pl ants was a bluishpink with white. Hacemes were fro m 2 1 to 26 cm long with about 47 flo wer sper raceme. Flowcring period per ra ceme varied from 10 10 14 da ys with3 to 4 fl owers open ed per racemc per da y.
D. intortum.: Flower color of D. int ortuni varied fr om pink to deep red, depcnding up on the plant. Racemes va ried from 10.8 to 14.6 em in len gth with27. 8 to 4:1 .4 f lowers per raceme. Flowerin g period per ra ceme vari ed fr om5.2 to 8 .4 days with 4.3 to 5.3 flowers opened per raceme per da y.
Pods and seeds: Th e three species studied produ ced ser ra ted pods with 4 to12 seeds per pod. Th c pods were covered with dense, uncin at e hai rs. Th e seedswere kidney sha ped, from 1.5 to over 3.0 mm long, and from 1.0 to over2.0 mm wide. Seed weight pe r 1000 seeds varied among the th ree species.
D. sandwiccnse : Percent age of pod se t fro m opcn polli nati on was 47.2 percent. Each pod pr odu ced an average of 6 .:1 seeds. Th e 1000- seed weigh twas 3.54 gra ms. Seed color was light brown. Seed germination was ab out99 percent for normal, matu re seeds.
D. uncinatum : P er centage of pod set fr om opcn pollination varied [rom16.9 to 32.6 per cent. Eac h pod produ ced about 4 seeds , Th c lOOO-sel'd weigh twas 4 grams. Seed color va ried from grcen to light brown. Seed germina tionwas about 96 percent for normal , mal ur e seeds .
D. intortum: Per cent age of pod set fr om open polli na tion var ied fr om 23.3to 51.1 percent. Seeds per pod va ried from 4.4 to 5.7. Th e 1000-seed weightwas 1.8 grams. Seed color was brown . Seed ge rmina tion was ab out 96 percent for normal, matu re see ds,
~ 0-
IT
AB
LE
A.Morpholo~ical
cha
rac
teri
stic
s,fl
ow
erin
gb
ehav
ior,
nu
mb
ero
fse
eds
per
po
d,
an
dp
erc
en
tag
ese
edg
erm
inat
ion
of
the
pare
nt
pla
nts
of
thre
eD
e,.m
otl
ium
spec
ies
5te
mch
arac
teri
stic
sIn
tern
ode
leng
thPa
rent
plan
tno
.C
olo
r(c
m)
Cro
ssse
ctio
nP
uhes
cenc
eG
row
thha
bit
Vig
or
D.
san
dlV
icen
se
511
......
......
......
......
......
......
......
......
......
..gr
een
3.2
roun
dfe
w,
shor
tup
righ
tfa
ir
521
......
......
......
......
......
......
......
......
......
..re
d3.
0ro
und
few
,sh
ort
upri
ght
fair
531
......
......
......
......
......
......
......
......
......
..gr
een
2.9
roun
dfe
w,
sho
rtup
righ
tfa
ir
D.
unci
nat
um
:I:U
I2...
......
......
......
......
......
......
......
......
.....
red
6.2
roun
dde
nse
,ho
oked
spre
ad
ing
good
> ::;;U
22...
......
......
......
......
......
......
......
......
.....
red
7.0
roun
dde
nse,
hook
edsp
read
ing
good
~U
32...
......
......
......
......
......
......
......
......
.....
red
6.6
roun
dde
nse,
hoo
ked
spre
adin
ggo
od> CO
l
U42
......
......
......
......
......
......
......
......
......
..re
d7.
5ro
und
dens
e,ho
oked
spre
adin
ggo
od::! o
U62
......
......
......
......
......
......
......
......
......
..re
d7.
3ro
und
dens
e,ho
oked
spre
adin
ggo
odc: r-
U82
......
......
......
......
......
......
......
......
......
..re
d7.
3ro
und
dens
e,ho
oked
spre
adin
ggo
od>
-i c::: :xl >
D.
into
rtu
mr- t'
l
113
......
......
......
......
......
......
......
......
......
....
red
4.1
obtu
sean
gled
dens
e,h
ook
edin
term
edia
tego
od>< ":
l:'
l12
3...
......
......
......
......
......
......
......
......
......
.br
own
3.4
obtu
sean
gled
dens
e,h
ooke
dsp
read
ing
fair
:::l i
133
......
......
......
......
......
......
......
......
......
....
gree
n4.
8ob
tuse
angl
edde
nse,
hook
edup
rig
ht
good
t'l :z
143
......
......
......
......
......
......
......
......
......
....
gree
n6.
3a c
ute
angl
edde
nse
,h
ook
edsp
read
ing
good
>-i
153
en...
......
......
......
......
......
......
......
......
......
.gr
een
4.7
acu
tean
gle
dde
nse,
hoo
ked
inte
rmed
iate
vigo
rous
>-i >
163
5.5
acu
tean
gled
few
,sh
ort
upr
igh
tgo
od>
-i...
......
......
......
......
......
......
......
......
......
..gr
een
0 :z
Infl
ore
scen
cech
arac
teri
stic
si
No.
ofF
low
erin
g10
00·
.., t'l
Rac
eme
flow
ers
peri
odFl
ower
sPo
ds
See
dsse
edSe
edee ~
Flow
erle
ngth
per
per
race
me
open
edse
t'pe
rw
eigh
tge
rmin
atio
nt'
lP
aren
tpl
ant
no.
colo
r'(e
m)
race
me
(day
s)pe
rda
y(p
erce
nt)
pod
(gra
ms)
(per
cen
t)0 :::J
D.
sand
wic
ense
'"' :IiS
l1...
......
......
......
......
...V
enet
ian
pin
k-
--
--
--
--< ee
S21
......
......
......
......
......
fusc
hine
pink
--
--
--
--
;:: "S3
1...
......
......
......
......
...da
wn
pin
k14
.950
.88.
85.
847
.26.
33.
5499
N >
D.
unci
nat
um
j 0 ~U
12...
......
......
......
......
..m
allo
wpu
rple
24.5
47.8
11.3
4.2
27.6
4.4
3.99
97>
U22
......
......
......
......
......
mal
low
pu
rple
4.05
rs:-
--
--
--
0
U32
......
......
......
......
......
mal
low
purp
le4.
0195
.5~
26.0
46.8
10.8
4.3
16.9
4.5
"U
42...
......
......
......
......
...m
allo
wpu
rple
21.2
48.0
12.0
4.0
21.3
4.1
4.08
-t;:
I '" '"U
62...
......
......
......
......
...m
allo
wpu
rple
22.6
46.0
14.0
3.3
32.6
3.9
4.11
96.5
5U
82...
......
......
......
......
...m
allo
wpu
rple
--
--
--
--
R.. ... ::::
D.
into
rtum
s en
113
......
......
......
......
......
phlo
xpi
nk"t
:-
--
--
--
-t'
l '"'12
3...
......
......
......
......
...ro
sein
epu
rple
10.8
27.8
5.2
5.3
23.3
5.7
--
;; en13
3...
......
......
......
......
...ph
lox
purp
le
143
......
......
......
......
......
phl
oxpi
nk14
.643
.48.
45.
251
.14.
41.
8296
.0
153
....
....
.....
....
.....
....
....
--
11.0
30.8
7.2
4.4
49.6
4.7
1.78
97.0
163
....
....
....
....
....
..._
_...
..--
--
--
-1.
9996
.5
'Flo
wer
colo
rac
cord
ing
toB
riti
shC
olou
rC
oun
cil,
Hor
ticu
ltur
alC
olou
rC
hart
,V
ol.
Ian
dII
.19
40.
Hen
ryS
tone
&S
on.
Ltd
.,E
ngla
nd.
,P
ods
set
und
erop
enpo
llin
atio
n.
-C
on
tin
ued
't>
:l--
-J
TA
BL
EA
.M
orp
ho
log
ical
char
acte
rist
ics,
flow
eri
ng
beh
avio
r,n
um
ber
of
seed
sp
er
po
d,
an
dp
erc
en
tag
ese
edI~
germ
inati
on
of
the
pare
nt
pla
nts
of
thre
eD
esm
od
ium
specie
s--C
on
tin
ued
Lea
flet
char
acte
rist
ics
Pu
bes
cenc
eS
ilve
rU
pper
Low
erL
engt
hR
atio
ofle
afle
tsu
rfac
esu
rfac
eX
wid
thle
ngt
hto
mar
kin
gP
aren
tpl
ant
no.
Col
orF
orm
Ape
xB
ase
(em
')(e
m')
(em
')w
idth
pres
ent
D.
sand
wic
ense
511
......
......
......
......
gree
nla
nce
olat
eac
ute
obtu
se72
1331
10.5
11.
74ye
s
521
......
......
......
......
gree
nla
nce
olat
eac
ute
obtu
se53
402
8.04
1.76
yes
531
......
......
......
......
gree
nla
nce
olat
eac
ute
obtu
se26
614
1312
.58
1.89
yes
D.i
nto
rtum
=:U
12...
......
......
......
...li
ght
gree
nel
lip
tic
acut
eob
tuse
135
258
26.9
61.
86ye
s> :;::
U22
......
......
......
......
ligh
tgr
een
elli
ptic
acut
eob
tuse
188
322
21.3
31.
80ye
s::
U32
......
......
......
......
ligh
tgr
een
elli
ptic
acut
eob
tuse
168
321
24.9
11.
84ye
s> o
U42
......
......
......
......
ligh
tgr
een
elli
ptic
acut
eo
btus
e17
728
222
.96
1.81
yes
a C"l
U62
......
......
......
......
ligh
tgr
een
elli
pti
cac
ute
obtu
se14
629
822
.63
1.79
yes
C r-U
82...
.....
......
......
....
ligh
tgr
een
elli
pti
cac
ute
obtu
se16
530
721
.66
1.77
yes
.., C :xl >
D.
into
rtum
t'"
Il3
1098
21.8
81.
62C
'l...
......
......
......
...gr
een
ovat
eac
ute
tru
nca
te25
7ye
sx ';
l
123
......
......
......
......
gre
enov
ate
acut
etr
unca
te18
540
518
.44
1.59
yes
C'l
:xl
133
......
......
......
......
gree
nov
ate
acum
inat
etr
unca
te11
197
32.0
11.
49no
i C'l
143
......
......
......
......
gree
nov
ate
acum
inat
etr
unca
te65
220
21.8
71.
58n
o:<: ..,
153
......
......
......
......
gree
nov
ate
acum
ina
tetr
un
cate
170
370
26.4
51.
56no
CIl ..,
163
acum
inat
e88
438
21.7
71.
55>
......
......
......
......
gree
nov
ate
tru
nca
teno
.., <5 :<:
UNIVERSITY OF HAWAIICOLLEGE OF TROPICAL AGRICULTURE
HAWAII AGRICULTURAL EXPERIMENT STATIONHONOLULU, HAWAII
HARLAN CLEVELANDPresident of the University
C. PEAIRS WILSONDean of the College and
Director of the Experiment Station
LESLIE D. SWINDALEAssociate Director of the Experiment Station