Plant Science 157 2000 13 – 22
Field evaluation of seed production, shattering, and dormancy in hybrid populations of transgenic rice Oryza sati6a and the weed,
red rice Oryza sati6a
James Oard
a,
, Marc Alan Cohn
b
, Steve Linscombe
c
, David Gealy
d
, Kenneth Gravois
e,1
a
Department of Agronomy, LSU Agricultural Center, Louisiana State Uni6ersity, Baton Rouge, LA
70803
, USA
b
Department of Plant Pathology and Crop Physiology, LSU Agricultural Center, Louisiana State Uni6ersity, Baton Rouge, LA
70803
, USA
c
Rice Research Station, LSU Agricultural Center, Crowley, LA
70527
, USA
d
Dale Bumpers National Rice Germplasm Center, Stuttgart, AR
72160
, USA
e
Uni6ersity of Arkansas Rice Research and Extension Center, Stuttgart, AR
72160
, USA Received 29 November 1999; received in revised form 25 February 2000; accepted 26 February 2000
Abstract
The genetic and agronomic consequences of transferring glufosinate Liberty™ herbicide resistance from transgenic rice Oryza sati6a L. lines to the noxious weed red rice Oryza sati6a L. were evaluated under field conditions. Replicated field trials in
Louisiana LA and Arkansas AR were conducted in 1997 to evaluate ten vegetative and reproductive traits of eight F
2
populations produced from controlled crosses of two transgenic, glufosinate-resistant rice lines and four red rice biotypes. Plant vigor and plant density at both locations were similar among populations derived from either transgenic or non-transgenic
parents. Significant differences in plant height and maturity were observed among LA populations produced from transgenic lines when compared to corresponding populations developed from non-transgenic material. However, values for these traits were not
greater than those detected in the red rice biotypes. Seed dormancy and seed production were not significantly different at either location among transgenic and non-transgenic populations. Dominant Mendelian segregation of glufosinate resistance was
detected in 40 of the populations evaluated. Results of this study indicated that those populations segregating for glufosinate resistance responded in a location-specific manner with respect to life history and fecundity traits. © 2000 Elsevier Science Ireland
Ltd. All rights reserved.
Keywords
:
Gene flow; Crop-weed hybrids; Fitness; Fecundity; Seed dormancy; Rice; Herbicide tolerance; Glufosinate; Liberty™ www.elsevier.comlocateplantsci
1. Introduction
Gene transfer technology in the last decade has produced various herbicide-resistant crop plants
[1] that are currently in or have the potential for commercial production. The BAR gene [2], iso-
lated from Streptomyces hygroscopicus, has been cloned and transferred to several crops [3 – 5] in-
cluding rice [6] for tolerance to the broad-spec- trum herbicide Liberty™. All trade names and
company names are listed for the benefit of the reader and do not imply endorsement or preferen-
tial treatment of the product by Louisiana State University, the University of Arkansas, or the US
Department of Agriculture. The active ingredient in this herbicide is the ammonium salt of glufosi-
nate ammonium-
DL
-homoalanine-4-yl methyl phosphinate, and will be referred to as glufosinate
in the remainder of this paper. The potential trans- fer of herbicide-resistant genes from transgenic
Approved for publication by the Director of the Louisiana Agric. Exp. Stn. as paper no. 00-09-0123.
Corresponding author. Tel.: + 1-504-3882110; fax: + 1-504- 3881403.
E-mail address
:
joardagctr.lsu.edu J. Oard.
1
Present address: Sugar Research Station, LSU Agricultural Cen- ter, PO Box 604, St. Gabriel, LA, 70776, USA.
0168-945200 - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 8 - 9 4 5 2 0 0 0 0 2 4 5 - 4
crops to weedy species and subsequent changes in fitness of weed populations have been a topic of
recent research. For example, interspecific F
1
hy- brids from Brassica rapa × transgenic Brassica na-
pus controlled crosses showed normal expression and Mendelian segregation of the BAR gene [7].
However, selfing of the hybrids produced no vi- able seeds, and transmission frequency of the BAR
gene was significantly reduced in one of two back- cross populations. Similar experiments using non-
transgenic material showed that B. rapa × B. napus F
1
, F
2,
and backcross generations were inter- mediate to or lower in fecundity than that of the
parents [8,9]. Transgenic B. napus × Raphanus raphanistrum wild radish F
1
hybrids and four successive generations were previously evaluated
under field conditions for seed production and BAR gene transmission [10]. Seed production was
low in F
1
populations, but increased to near wild type levels by the third generation. BAR frequency
decreased to 25 of the original by the fourth generation and was associated with reduced num-
bers of chromosomes in female parents. Hybrids produced between wild and domesticated strains
of sunflower Helianthus annus L. showed reduced seed production and dormancy for most, but not
all, of the F
1
populations that were evaluated [11]. Wild strains and domesticated cultivars of rice
have been shown to hybridize and produce viable offspring, but with varying degrees of efficiency
[12]. Cultivated rice has a companion weed, red rice Oryza sati6a, which exists in temperate and
tropical regions of the world that reduces grain yield and quality and is considered a noxious weed
in rice producing areas of the southern United States [13]. At present, red rice control is achieved
through crop rotation and paddy-water manage- ment. Frequency of hybridization between non-
transgenic cultivated rice and red rice has been evaluated in one study [14], where seeds were
collected from red rice plants found in commercial Louisiana rice fields and planted in common gar-
den experiments. Electrophoretic analysis iden- tified individuals produced from hybridization
between cultivated and red rice plants. Hybrids generally exhibited greater height and flag leaf
area than the cultivated or red rice parents. Tiller number was greater in the hybrids than in the red
rice, but not in the commercial cultivars. Overall, these results indicated that hybrid vigor did occur
for certain vegetative characteristics. Incidence of hybridization ranged from 1 for early maturing
cultivars to a high of 52 in a commercial plant- ing of the late maturing cultivar Nortai.
The BAR gene was recently transferred and evaluated in 11 different transgenic rice lines in 2
years of field-plot trials [6]. Significant differences among transgenic BAR-containing lines were ob-
served for grain yield, plant height, and date of flowering before or after treatment with Liberty.
Other work [15] showed that insertion of the nptII gene in rice was associated with reduced seed
fertility, delayed maturity, and smaller flag leaves when compared with the corresponding non-trans-
formed, protoplast-derived plants. Similar results were obtained in transgenic barley [16]. Reciprocal
crosses between two transgenic rice cultivars and a common Louisiana red rice biotype displayed sin-
gle-gene, dominant transmission and expression of the BAR gene in F
1
and F
2
generations [17]. No cytoplasmic influence on expression of the trans-
gene was detected when either the cultivated or wild strains were used as maternal parent. A total
of five quantitative trait loci QTL were detected for rice seed dormancy on four chromosomes that
explained 48 of total phenotypic variation in BC
1
F
5
lines [18]. One dormancy QTL mapped to the same location as a QTL for heading date.
The objective of this research was to investigate agronomic and fitness traits of BAR transgenic
rice-red rice hybrid populations that segregated for resistance to Liberty herbicide under field condi-
tions. We expected to gain a greater understanding of potential consequences of gene flow from herbi-
cide resistant commercial rice into red rice. To our knowledge this study was the first in the United
States to evaluate potential effects of the BAR gene on life history, fitness and seed characteristics
of red rice.
2. Materials and methods
