|
|
||||||||
Tropical Forages Project, International Center for Tropical Agriculture, A.A. 6713, Cali, Colombia
* Corresponding author (j.miles{at}cgiar.org)
| ABSTRACT |
|---|
|
|
|---|
| INTRODUCTION |
|---|
|
|
|---|
Commercialization of brachiariagrasses in Latin America began in the early 1970s with the importation to Brazil of Australian seed of Basilisk signalgrass (Santos Filho, 1996), a cultivar derived directly from a Ugandan germplasm accession introduced to Australia in 1930 (Oram, 1990). Adoption of cv. Basilisk was very rapid in Brazil and more widely throughout tropical America as it proved to be well adapted to extensive areas of savanna with low fertility and low pH soils (Santos Filho, 1996).
The greatest biotic limitation on brachiariagrass pasture productivity in the neotropics are the many native genera and species of grassland spittlebugs (Homoptera: Cercopidae) (Valério et al., 1996; Holmann and Peck, 2002). Basilisk signalgrass and common commercial ruzigrass are particularly susceptible. Marandu palisadegrass was released by the Brazilian Agricultural Research Corporation (EMBRAPA, its Portuguese acronym) in the mid-1980s owing mainly to its resistance to spittlebugs (Nunes et al., 1984). Marandu is now extensively sown, though its adaptation does not extend to the very low fertility soils where cv. Basilisk is so well adapted, nor is it used in parts of the tropics (e.g., Thailand or tropical Australia) where spittlebug attack is not a threat. Ruzigrass, though of high nutritional quality (Valle et al., 1988; Lascano and Euclides, 1996), is only a minor commercial species in tropical America owing to its poor adaptation to low fertility soils and particularly to its extreme susceptibility to spittlebugs (Keller-Grein et al., 1996). Ruzigrass is an important commercial forage grass in southeast Asia, e.g., in Thailand, where locally produced seed is abundant and cheap (Hare and Phaikaew, 1999).
A small collection of brachiariagrass germplasm maintained by the International Center for Tropical Agriculture (CIAT, its Spanish acronym) was augmented during 1984 and 1985 through direct collection in Africa (Keller-Grein et al., 1996) with the objective of obtaining genotypes with broad edaphic adaptation combined with resistance to spittlebugs. Agronomic evaluation of the new germplasm collection suggested that none of the accessions offered the needed combination of attributes. Following the development of a fully sexual, tetraploid ruzigrass biotype (Swenne et al., 1981) and confirmation of its cross compatibility with apomictic tetraploid signalgrass and palisadegrass accessions (Ndikumana, 1985), applied brachiariagrass breeding programs quickly ensued, both at EMBRAPA and at CIAT (Miles and Valle, 1996; Miles et al., 2004).
Inheritance studies indicated a single-locus model of genetic control of apomixis in the signalgrasspalisadegrassruzigrass agamic complex (Ndikumana, 1985; Valle et al., 1994; Valle and Savidan, 1996): tetraploid apomicts are simplex for a dominant allele (Aaaa) at the putative "apomixis locus," sexuals being homozygous recessive at the locus and either diploid (aa) or tetraploid (aaaa). Hence, tetraploid, sexual x apomictic hybrid progenies segregate approx. 1:1 for reproductive mode and sexual x sexual progenies are uniformly sexual (Valle and Savidan, 1996).
While the tetraploidized sexual ruzigrass provided a means to achieve genetic recombination, this germplasm has a very narrow genetic base (J. Tohme, unpublished molecular marker data) and is both poorly adapted to infertile acid soils (Rao et al., 1998) and highly susceptible to spittlebugs (Lapointe et al., 1992; Cardona et al., 2004). Since sexual ruzi-, signal-, and palisadegrass and their hybrids have been shown to be allogamous and self-incompatible (Ngendahayo et al., 1988), we adopted a strategy to synthesize a broad-based, tetraploid sexual population by intercrossing sexual clones from sexual x apomictic hybrid progenies, followed by recurrent selection on spittlebug resistance and other traits (Miles and Valle, 1996; Miles and Escandón, 1997; Miles et al., 2004). Given that sexuality is conditioned by the homozygous condition at the apomixis locus, a sexual population advanced by open pollination will remain fully sexual if not contaminated with pollen from compatible apomicts (Miles and Escandón, 1997). Genetic gain in the sexual tetraploid population would be captured in apomictic cultivars by crossing the best sexual clones each cycle with one or more elite apomicts, followed by isolation of superior apomictic segregates from the resulting hybrid populations.
Initial emphasis in improving the synthetic sexual population was on spittlebug resistance. Miles et al. (1995) presented data suggesting that spittlebug resistance in the brachiariagrasses may be under relatively simple genetic control and highly heritable. They suggested recurrent selection in a sexually recombining population as a means to enhance expression of resistance. A reliable, but cumbersome, bioassay based on artificial infestation of pot-grown plants was available when the breeding program began (Lapointe et al., 1989a, 1989b). Subsequent refinements in methodology greatly improved throughput and allow the separate assessment of both tolerance and antibiosis (Cardona et al., 1999). A modification in screening protocols came with the appreciation of interaction of host genotype with spittlebug species (Valério et al., 2001; Cardona et al., 2004). This interaction suggests the prudence of testing against all important pest species.
The results of the artificially infested, greenhouse spittlebug bioassay have been repeatedly confirmed in field trials in the Colombian Amazonian piedmont where A. varia and Z. carbonaria are prevalent and natural spittlebug pressure is high (CIAT, 2004).
The present report describes an ongoing, applied brachiariagrass breeding program. Its main objective is to document improvements in resistance to three Colombian spittlebug species resulting from six cycles of recurrent selection in a broad-based, synthetic, sexual tetraploid breeding population.
| MATERIALS AND METHODS |
|---|
|
|
|---|
From C2 to C6, the synthetic tetraploid sexual population was advanced on a 2-yr cycle. Large seedling populations (2588 to 4343 individuals) were culled on general agronomic performance (vigor, flowering, growth habit, leafiness, healthy foliage) assessed in randomized, unreplicated, space-planted field trials at each of two contrasting locations in Colombia during Yr 1. In Yr 2, clones surviving the initial cull were tested for spittlebug reaction and final selections were recombined by open pollination in an isolated crossing block.
Spittlebug Bioassay
Preselected clones (those surviving the initial cull on field performance) were vegetatively propagated to CIAT headquarters and established in 25.4-cm pots. Single-tiller propagules were taken from these plants to establish the spittlebug experiments under greenhouse conditions [mean temperature 24°C (range: 1927°C); mean relative humidity 75% (range: 7090%)]. Both susceptible and resistant checks were included in every experiment. The protocols for the bioassay differed from cycle to cycle (Details given in Table 1) as methodologies evolved over time (Lapointe et al., 1989a; Cardona et al., 1999). In general, individual experimental units were infested with a uniform number of mature spittlebug eggs (Table 1). Plant damage caused by nymphal feeding was rated on a 1-to-5 visual scale (1 = no detectable damage; 5 = dead plant; half-unit scores [1.5, 2.5,..., 4.5] were also used for a total of nine classes). Normal surviving insects were counted either at the final (fifth) nymphal instar or as emerging adults. Survival counts were made for all genotypes tested in C2, but in later cycles were conducted only for checks and for plants with a low damage rating (
2).
|
A two-phase screening procedure was implemented in C3 and subsequent cycles. In the first phase, several hundred (range: 502928) preselected genotypes were tested with minimal (1 or 2) replication. Following culling on this first phase (8490% culled: Table 1), resistance was reconfirmed in a second phase in replicated trials and the most resistant clones identified for recombination. These final selections were made attempting to take into account both low damage score and low survival for all three spittlebug species but without recourse to a formal selection index. The number of final selections recombined ranged over cycles from 11 to 42 (Table 1).
In the most recent selection cycle (C6), 756 preselections were screened between January and March 2004, against the three Colombian spittlebug species. In this first, unreplicated phase, 120 genotypes showed no plant damage and no survival of nymphs for all of the three insect species. The subset of 120 apparently resistant clones was retested between April and July 2004, in three replicated trials, one for each spittlebug species. These trials included the 10 ancestral accessions that were involved in the synthesis of the original sexual tetraploid population (Table 2). Resistance of the set of 120 sexual hybrid clones was reconfirmed in three replicated experiments conducted between October 2004 and January 2005, with standard checks.
|
Analysis of variance was performed on data from the replicated entries for each of the spittlebug trials. The residual mean square was taken as an estimate of error variance for each trial. Means were compared by t test with a comparisonwise error rate of
= 0.05, or 0.01.
| RESULTS |
|---|
|
|
|---|
The four most spittlebug-resistant ancestral parental genotypes constituted just over one-quarter (25.86%) of the C0 gene pool, the remaining three-quarters being comprised of susceptible accessions. Hence, the early cycles of the synthetic tetraploid sexual population were expected to be susceptible. Indeed, C2 preselections were as susceptible (to A. varia) as the susceptible tetraploid ruzigrass check (mean survival of 72.3 vs. 69.0% for the preselections or ruzigrass, respectively). Even the most resistant, final selections in C2 were more susceptible than Marandu to A. varia (Table 3).
|
The mean resistance of the subset of clones selected each cycle for recombination has improved consistently relative to cv. Marandu (Table 3). Mean resistance (to A. varia) of the set of final selections superior to that of Marandu was not achieved until C5. In C6, clones selected for recombination were more resistant to all three species than Marandu (Table 3).
In C5, for the first time a single clone more resistant than cv. Marandu to all three spittlebug species was identified among approximately 800 preselections. This clone (SX01NO/0102) has since been adopted as a resistant check and its outstanding combined resistance fully confirmed (CIAT, 2004).
The 120 clones retested for spittlebug reaction in C6 represent the upper tail of the C6 population distribution. They are more resistant as a group than cv. Marandu [mean survival in the reconfirmation trial of 13.0 vs. 36.7%; 2.6 vs. 52.4%; or 6.6 vs. 50%, for A. varia, A. reducta, or Z. carbonaria, respectively (means in each pair differ, by t test, at the
= 0.01 probability level)]. On 15 of the 120 clones, nymphs of all three spittlebug species failed completely to survive. These 15 clones are at least as resistant as the single highly resistant sexual tetraploid clone identified in C5 (SX01NO/0102: 0.0, 3.3, or 0.0% survival of A. varia, A. reducta, or Z. carbonaria nymphs, respectively). SX01NO/0102, in turn, is as resistant or more resistant to each of the spittlebug species as the most resistant of the ancestral parental clones (Table 2).
| DISCUSSION |
|---|
|
|
|---|
It is not clear why unimproved African grasses, such as cv. Marandu, should exhibit resistance to neotropical insect pests, since natural selection in Africa would not specifically have elicited this resistance. Apparently some nonspecific, general resistance factors are involved, factors that coincidentally confer resistance to American spittlebugs. Our results confirm the existence of alleles that negatively affect spittlebug nymphal development in these African grasses, though the physical or biochemical basis of the observed resistance is unknown.
Miles et al. (1995) concluded that spittlebug resistance in brachiariagrasses was "probably not complex genetically, although more than a single major resistance gene are likely involved." The results reported here are entirely consistent with this statement: While the present study does not allow quantification of selection response, our data do demonstrate that resistance improves in response to intense, recurrent selection on reliable phenotypic data. No evidence of any major resistance gene effect was detected.
Both tolerance and antibiosis mechanisms of spittlebug resistance are documented for brachiariagrasses (Cardona et al., 2004). Since brachiariagrasses are perennials, grown on large continuous areas, spittlebug population increase on a tolerant cultivar can lead to resistance being overcome (Ferrufino and Lapointe, 1989; Lapointe et al., 1992; Valério et al., 2001; and Cardona et al., 2004). Hence, the aim of the breeding program is to develop cultivars with true antibiosis resistance. Our results demonstrate that antibiosis resistance increases in response to selection and that resistance to several spittlebug species can be combined.
The eventual durability of the spittlebug resistance being developed in the synthetic breeding population remains to be determined. It is encouraging that more than two decades after cv. Marandu was released, its resistance (Nilakhe, 1987) remains effective in the field throughout the American tropics where Marandu is exposed to a diversity of pest spittlebug species over massive areas of essential monoculture. Preliminary information on intraspecific geographic diversity in insect aggressiveness is encouraging, since no interaction with host genotype has so far been detected (CIAT, 2004).
Concern is sometimes expressed of possible negative effects of insect resistance factors on the feeding quality of forage species (e.g., Sorensen et al., 1988). While it is necessary to remain alert to this possibility, we know of no evidence to support such a correlation in the brachiariagrasses: the resistant cv. Marandu produces equally high quality forage as does the susceptible cv. Basilisk (Euclides et al., 1993; Lascano and Euclides, 1996) and a spittlebug-resistant hybrid, CIAT 36087, has even better quality.
The identification of highly spittlebug-resistant, sexual tetraploid clones represents an important advance in Brachiaria breeding. However, the resistance achieved by selection in the sexual population will have no direct commercial value until it is captured in apomictic clones that can be released as true-breeding cultivars. We anticipate, but have not yet demonstrated, that highly resistant sexual clones will consistently produce sexual x apomictic hybrids with useful levels of resistance. Sanford and Ladd (1987), for example, found that approximately half of the genetic gain realized from recurrent selection in potato (Solanum tuberosum L.) for resistance to potato leafhopper [Empoasca fabae (Harris)] was expressed in hybrids between the resistant population and susceptible clones.
The very first hybrids between highly resistant sexual clones and susceptible apomicts (e.g., cv. Basilisk) have only recently become available so the expression of resistance in these hybrids is still unknown. Nor have the best, multiple-species-resistant clones been exposed to non-Colombian spittlebug species to determine how broadly effective their resistance is.
When it is confirmed that adequate spittlebug resistance has been achieved in our synthetic tetraploid sexual population, then greater attention in the population improvement program will need to be focused on other attributes required in commercial brachiariagrass cultivars, for example resistance to foliar blight caused by Rhizoctonia solani Kühn and high feed quality.
| ACKNOWLEDGMENTS |
|---|
Received for publication June 3, 2005.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. W. Miles Apomixis for Cultivar Development in Tropical Forage Grasses Crop Sci., December 18, 2007; 47(Supplement_3): S-238 - S-249. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||