NATURAL HYBRIDIZATION AMONG THREE SYMPATRIC BAPTISIA (FABACEAE) SPECIES IN NORTH CENTRAL TEXAS

22 Pages·1996·12 MB·English

by  Matthew A Kosnik

AMONG
NATURAL HYBRIDIZATION THREE
SYMPATRIC
(FABACEAE)
BAPTISIA SPECIES
NORTH
CENTRAL
TEXAS
IN
Department
Biology
of
Austin
College
TX
75090-4440
Sherman, U.S.A.
BARNEY
LIPSCOMB
L.
Botanical Research Texas
Institute of
309
Pecan
Street
TX
76102-4060
Fort Worth, U.S.A.
A
com
north central Texas Baptism (Fabaceae, Faboideae, Thermopsideae) hybrid
hybrids investigated using morphological, spectrophotometric, and isozyme ela
is
£
Morp
species and all the possib le Fi hybrids. >hological ^idence for mt,regressive hy
alis.^dB.sphaerocarpa
preseinted.
IS
RESUMEN
Un
complejo de hfb ndos del norte d Texas, B\aptista (Faba.ceae, Faboi^
e
Thermopsideae) que conti,enetres padres (6.,australis, B. sphaeroatrpa, B. bra cteata)
y
ydee
hibndosFiseinvestigapo metodos morfolog espectirofotometricos, lectrof
:icos,
ir
x nomb
de isozimas. Dos de los res hibndos (B. bushii y B. < bkolor] habi'a
t ;
/ ) 1
X xvar
Mrpa
previamente, pero el hibr ido de B. austrails B. sphaero. (B iicolor se des
)
.
mo
Baptism y todos los hibi Se preset .idenci a rfologi
,.
ZarpI
y B, sphae
'is
The commonly Wild
genus Baptista, called Indigo, a perennial, her-
is
25-30
aceous, and conspicuously flowered legume genus with about spe-
occurring in the eastern United States. In Texas they occur in remnant
[es
common
ative habitats and are weeds in overgrazed pastures (Turner 1959)-
common
Taxonomic have been the group due extensive
difficulties in to
&
among
and
hybridization introgression species (Correll Johnston 1970).
This taxonomic confusion and the proliferation of secondary compounds
within the genus led to early use of chemical taxonomic methods
(e.g.
& &
chromatography by Turner Alston 1959; Alston Turner 1962; Alston
&
Turner and
1963), later detailed investigation of flavonoid distributions
&
(Markham Dement Mabry
within the genus et 1970; 1975). Because
al.
known
of this work, a great deal about Baptism both morphologically
is
and
chemically.
Three
Baptisia species occur in the area of north central Texas studied, B.
australis (L.) R. Br. ex W.T. Alton var. minor (Lehm.) Fern, (a blue-violet-
flowered species, Fig. B. sphaerocarpa Nutt. yellow-flowered
(a species,
1),
&
and Muhl.
Fig. B. bracteata ex kucophaea (Nutt.) Kartesz
Ell. var.
1),
Gandhi cream-flowered Nomenclature
(a species, Fig. follows Kartesz
1).
(1994). Baptisia australis reaches the southern extreme of range in the
its
north central Texas area, B. bracteata reaches western boundary in the
its
area, and B. sphaerocarpa reaches northern extreme in central Oklahoma
its
The
where
(Larisey 1940a). area three ranges overlap limited north
to
all
is
central Texas, southern Oklahoma, and western Arkansas
(Fig.
2).
Hybridization and multispecies hybrid complexes are well documented
in the genus Baptisia, both through traditional morphological methods
and through extensive chemical analysis. In her monograph on the genus
Baptisia, Larisey (1940a) recognized eight Baptisia hybrids and cited the
relatively frequent occurrence of hybridism, particularly in areas where the
At
borders of the geographically larger species overlap. several north cen-
Texas locations in Grayson County and neighboring Fannin County,
tral
several Baptisia species coexist and have been suspected of hybridizing
X
The
(Kosnik 1996). B. australis B. bracteata hybrid, with an intermediate
x
or light blue-violet flower (Fig. has been described previously
{B.
1),
& X
Greenm.
and
bicolor Larisey), the B. bracteata B. sphaerocarpa hybrid,
with an intermediate or light yellow flower (Fig. has also been described
1),
X X
previously {B. bushii Small). However, the proposed B. australis B.
sphaerocarpa hybrid, with a flower color varying from brick-red and yellow
to orangish to bluish-violet and yellow (Figs. and has not been previ-
1 3),
ously described in the literature.
and
Baptisia bracteata, B. australis, B. sphaerocarpa have also each inde-
members
pendently been observed hybridizing with other of the genus.
X
Baptisia australis B. bracteata hybrids have been observed in Kansas
X
(Hitchcock 1894). Baptisia sphaerocarpa B. laevicaulis hybrids have been de-
scribed using morphological and chemical methods in southeastern Texas by
Turner and Alston (1959) and Alston and Turner 962). Baptisia sphaerocarpa
( 1
X
B. leucantha hybrids were observed in Oklahoma by Engelmann (1878),

iio. 2. Di^Li.buLiui. uiap for the piirencal species. (Modified from Larisey i940a).
and
in southeastern Texas they were described morphologically by Larisey
(1940b) and morphologically and chemically by Alston and Turner
(1962).
x
Bapthia
sphaerocarpa B. bracteata hybrids were observed in Louisiana by
Harper However,
(1938). only in north central Texas that three of
it is all
and
these species three possible F] hybrids have been observed mixed
all in
populations. Introgressive hybridization (Anderson move-
1949), or the
ment
of genes from one species into another via hybrids, has
fertile also
&
&
been suggested for Baptisia (e.g., Alston Turner 1963; Correll Johnson
1970).
METHODS
—
Collection Plant material was collected 11 Grayson
localities. at sites in
and Fannin At
counties, Texas. three additional hybrids were observed
sites
but not collected. Table summarizes collections and observations. Coor-
1
dinates for each site (Tible were determined using a Garmin 45 Global
1)
Positioning System. Voucher specimens
are deposited the Botanical Re-
at
search Institute of Texas (BRIT) Fort Worth.
in
—
Morphological Measnrements. Measurements of a number of characters were
made
in the field, prior to collection and pressing. Vegetative height was
measured from ground cm
the to the highest leaves (0.5 increments). Plant
height was measured from ground maximum cm
the to the height (0.5
Raceme
increments). angle was measured with weighted
protractor (2.5°
a
0°
increments, considered vertical and 90 considered horizontal). Total flower
length (from the base of the flower to the tip of the keel or wings), total
mf\Wf^^^^^^
^^j^gr-^l^^mmmm
1
'y>«
GEO-B
HAG-C
MEM-CY
flower height (from the bottom of the keel or wings to the tip of the ban-
maximum
maximum
ner), horizontal width of the wings, and banner width
mm
were measured with
digital calipers (0.01 increments). All flower
measurements were made on the lowest fully developed, intact flower on
the best developed inflorescence, with at least one fully developed flower
Measurements made
of other vegetative were
all characters after press-
mm
ing using digital calipers (0.01 increments). These included pedicel
maximum
length (of the measured flower); bract length, width, width
floral
maximum
attachment;
at petiole length; central length, width,
leaflet dis-
maximum
tance to the width (from the leaflet base); and stipule length,
The
(from the stipule base). leaflet measurements were taken on the center
undamaged,
leaflet of the first well developed leaf below the raceme. Stipule
measurements were taken from the same leaf, unless the leaf did not have a
which
stipule, in case the stipule closest to that leaf was measured (this was
with
often the case B, sphaerocarpa).
Fruit characters were chosen to examine for evidence of introgressive
hybridization between B. australis and B. sphaerocarpa because of the very
different fruit shapes and sizes of these taxa and thus the increased prob-
unambiguous
ability of finding evidence of introgression. Individuals from
the hybrid population were labeled as to flower color during the spring
when
flowering season and fruits were subsequently collected mature. Fruit
mm
measurements were taken using increments) from
digital calipers (0.01
FAN-
completely dried fruits collected in the in June and July from
field
PWR-B
GEO-B, HAG-Y, maximum
A6, and (Table These included
1).
fruit length (not counting stipe or beak), height, width, and length of beak
(from where fruit body narrows dramatically apically to tip of beak). Mea-
surements were taken on the lowest well developed, well preserved on
fruit
Colour Chart (1966) was used
also in
'
The
easurements. lowest, mature, un-
damaged on raceme was matched
flower the color before dried.
it
—
Morphological measurements were analyzed using
Statistical Analyses.
component
principal analysis, multiple discriminant analysis multivari-
(a
ate statistical technique sometimes called canonical analysis), and analysis
of variance and run with Minitab 10.5Xtra (1995) and BioStat (Pimentel
II
1995) on a Macintosh Centris 650. Because missing data are not allowed in
and were
these analyses, only the characters listed in Tables 2 3 used. Prin-
component
cipal analysis attempts to unravel patterns of variation within
the entire dataset, while multiple discriminant analysis attempts to mini-
mize within group while maximizing between group
variation variation.
Discriminant computes by weighting
analysis synthetic characters dataset
characters to maximally separate the groups. For the discriminant analysis
individuals were assigned one of groups based on flower
to six color.
—
Spectrophotometric Analysis. Flowers from four to nine individuals of each
of the taxa were examined spectrophotometrically. Flowers from three
six
MEM-CY,
(FAN-A6, PRE-A6)
and were selected to represent the
sites
maximum
variation seen within the populations. Petals from a fresh, un-
1%
damaged, mature flower were ground completely in 2 ml of a 1:4, HCl:
EtOH
100%
solution; the samples were then centnfuged for three min-
One ml
utes to separate large particulate material. of the resulting extract
An
ml
was diluted with an additional 2.5 of solvent. ultraviolet-visible
DU-70
Beckman
absorbance spectrum was then measured using
light a
scanning spectrophotometer; absorbance scans were run 1200 nm/min.
at
nm
from 200 700 nm. Absorbances peaks 422 nm, 449 nm, and
to for at
nm
nm)
540 (±2 were recorded.
—
Enzyme Leaf samples were obtained from 21
Electrophoretic Analysis. in-
dividuals including B. australis, B. sphaerocarpa, and all three F] hybrids at
FAN-A6
in July 1990. Baptisia bracteata was not examined. Starch gel
enzyme
electrophoretic procedures generally followed Soltis et (19H3).
al.
The
tris-HCl grinding buffer-PVP solution described by Soltis et (1983)
al.
PVP
12%
was used with a 40,000 concentration of (wt/vol). Starch gel
(AAT
The following enzymes were resolved: aspartate aminotransferase
GOT), (G3PDH),
glyceraldehyde 3-phosphate dehydrogenase
or isocitrate
(MDH),
dehydrogenase (IDH), malate dehydrogenase 6-phosphogluconate
dehydrogenase (6PGD), shikimate dehydrogenase (SKDH), and
triosephosphate isomerase (TPI). Buffer systems follow Soltis et (1983):
al.
G3PDH SKDH
and were resolved on a modification of gel and electrode
GOT
TPI
buffer and on a modification of buffer 8 (modification de-
1;
& 6PGD
scribed by Soltis Soltis 1987); on a modification of buffer and
9;
MDH M
IDH
and on morpholine system 0.04
a (electrode buffer: citric
pH
acid, 7.5 by addition of N-(3-aminopropyl)-morpholine; gel buffer:
dilute electrode buffer 1:25). Staining procedures followed Soltis et
al.
(1983).
—
Based simply on
Morphological Analysis. flower color, the three parents
A
and their intermediate putative hybrids are easily distinguished. variety
of morphological characters also suggest hybridization. Simple two charac-
ter plots of raceme angle versus flower size (height) or leaflet size (length)
distinguish six basic morphological groups which support the hybridiza-
The
mor-
tion hypothesis (Fig. 4). three hybrids are generally intermediate
phologically between the parents. These graphs also show considerable
among
overlap hybrid
taxa.
component manner
Principal analysis separates the taxa in a similar to
two The component
the character plots. plot of the principal versus the
first
component
second principal illustrates that variation within all six taxa (3
parental species and 3 hybrids) less than the variation between taxa (Fig.
is
—
5) in other words, groups can be observed though there some overlap.
is
Individuals of the three parental species form discernable clusters toward
the periphery of the plot, while the suspected hybrid taxa are generally
all
intermediate between the purported parental species. Character weightings
given Table
are in
2.
Discriminant analysis separates the taxa with greater resolution than the
component
principal analysis, but does not completely separate
still all
94%
The
individuals of the taxa (Fig. discriminant analysis supports
all 6).
6%
The which
of the original flower color grouping. of the individuals are
ambiguously grouped by the discriminant analysis on the basis of the mor-
phological data are clearly distinct in other characters, such as flower color.
The
linear discriminant functions are given in Table
3-
addit
I
alysis
and These
because they could not be accurately consistently quantified.
numerous on
include hairs (too to count effectively B. bracteata), flower
maximum maximum
bracts (length, width, and distance to width cannot
be measured on plants without persistent bracts), and flower color (not
contmuously quantifiable). Bapttsia bracteata densely pubescent, has large
is
mm
(10-30 and cream while
persistent bracts long), flowers, B.
floral
sphaerocarpa and B. australis are glabrous, have smaller, early deciduous bracts
mm
(0-5 long), and have vividly colored flowers (bright yellow in the case
The
of S, sphaerocarpa or blue-violet in the case of B. australis). hybrids, B.
X X X
and
bushii {B. sphaerocarpa B. bracteata) B. bicolor {B. australis B.
4- ^i\;<\
^A^O<^
1
'¥^*
^o^
1
1-
>
'o'
-
. .
Yellow
'(Blue X
Component
Principal 2
between
bracteata), are intermediate their respective parents in of these
all
characters. Baptisia sphaerocarpa and B. amtralh are easily distinguished by
flower color and plant form. Baptisia sphaerocarpa has numerous
flexible,
commonly
vegetative branches creating a bushy form, multiple racemes
much
per plant, and intense yellow flowers. Baptisia australis has fewer,
more
rigid branches, typically only one raceme per plant, and a blue-violet
flower color that varies greatly in saturation between plants. Baptisia variicolor
more
generally vegetatively intermediate, but often superficially simi-
is is
lar to B. sphaerocarpa; they have vegetative branches creating a bushy form
X
and often have multiple racemes per plant. The flower color of B. variicolor
is strikingly variable. It ranges from intermediate to a patch-work combi-
nation of the parental flower colors, with different hybrid individuals hav-
ing flowers ranging from brick-red and yellow to orangish to bluish-violet
and yellow. This variation can be seen in Figs. and The Royal Horti-
1 3.
cultural Society Colour Chart measurements are an indication of the varia-
tion in color (Table
4).
—
The
absorbance
Spectrophotometric Analysis. spectra indicate three diag-
wavelengths peak 540 nm,
nostic (Fig. 7). Baptisia australis has a large at B.
nm
sphaerocarpa has distinct peaks at 422 and 449 nm, and B. bracteata has
low absorbance at all of the diagnostic wavelengths. Baptisia variicolor clearly
x X
shows both parental peaks, while and have peak
B. bicolor B. bushii