Published in Crop Sci. 44:1291-1298 (2004).
© 2004 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
CROP ECOLOGY, MANAGEMENT & QUALITY
Indehiscence Expression and Capsule Anatomy in Vernola
F. Verdolinia,
A. Anconetania,
D. Lauretib and
M. J. Pascual-Villalobosc,*
a Instituto Murciano de Investigación y Desarrollo Agrario y Alimentario, Estación Sericícola, 30150 La Alberca, Murcia, Spain and Istituto Sperimentale per le Colture Industriali, 60027 Osimo (AN), Italy
b Istituto Sperimentale per le Colture Industriali, 60027 Osimo (AN), Italy
c Instituto Murciano de Investigación y Desarrollo Agrario y Alimentario, Estación Sericícola, 30150 La Alberca, Murcia, Spain
* Corresponding author (mjesus.pascual{at}carm.es).
 |
ABSTRACT
|
|---|
Euphorbia lagascae Spreng. is a potential new industrial crop because of the synthesis of vernolic acid in about 65% of the total fatty acids in its seed oil. Breeding research in the 1990s gave rise to the indehiscent Vernola cultivar as opposed to a complete dehiscence of the wild types. With the aim to evaluate the environmental influence in the expression of the nonshattering trait, field trials were sown in spring and autumn in Murcia, Spain, and Ancona, Italy, during 2001 and 2002. To further examine the basis of partial indehiscence, a capsule wall histological study was also performed. There was a significant decrease (P < 0.01) from 86.9 to 67.9% in the number of capsules retained per plant in the spring sowing, when mister sprinklers were used to increase relative humidity (RH) during pod formation. Overall, 37.9% indehiscent capsules per plant were retained in the autumn cycle vs. 77.4% in the spring cycle, indicating a strong influence of the growing season, which was associated to a combined effect of higher moisture and lower temperatures during capsule development. Transversal sections of carpel walls from Vernola capsules showed that the physical reason of the partial dehiscence observed in this cultivar is a small portion of the mesocarp remaining near the valve sutures, which makes the fruits susceptible to a slower opening under some environmental conditions.
Abbreviations: RH, relative humidity
|
INTRODUCTION
|
|---|
EUPHORBIA LAGASCAE (Euphorbiaceae) is a potential new industrial crop (Turley et al., 2000). It has a seed oil content ranging from 48 to 52%, with vernolic acid comprising between 58% and 67% of total oil content. The oil is of significant interest in lubricant and polymer industries as a biodegradable replacement of petrochemical oils or as an oil with new properties.
Indehiscent genotypes were developed cooperatively by Centro de Investigación y Desarrollo Agroalimentario (Murcia, Spain) and Institut für Pflanzenbau und Pflanzenzüchtung (Göttingen, Germany). Germplasm was registered (Reg. no. GP-25, PI 604649) through the USDA (Pascual-Villalobos, 1999). Cultivar Vernola has variety protection (No. EUG2) after being tested by the Bundessortnamt (Hannover, Germany) in 1998. The characteristic of this cultivar is capsule indehiscence as opposed to a complete dehiscence of the wild E. lagascae types.
In early field experiments in Spain, it was observed that retention of capsules in nonshattering genotypes was on average 72.7%, with fall plantings less favorable for the expression of this character (Pascual-Villalobos et al., 1994). The authors reported thinner fruit wall and lack of a pericarp layer in nonshattering capsules. Fahn and Zohary (1955) stated that the absence of a sclerenchymateus stratum in legume pods was responsible for the indehiscence character. Domestication of many cereals, legumes, and oilseeds has been linked to the selection of nonshattering types.
Euphorbia lagascae has several undesirable characteristics, namely indeterminate growth, a long maturation period, the presence of latex, and susceptibility to weed competition. A 4-yr evaluation in southern England demonstrated that E. lagascae has potential for cultivation, with seed yields of about 1.1 Mg ha–1 on average, and no major limitations from the dehiscent nature of the wild types (Cromack, 1999).
Perez Marco (1999) observed good adaptation in dry areas of northeast Spain and suggested that autumn plantings would make better use of winter rainfall and that increasing plant densities resulted in higher yields.
Seed dehiscence causes yield losses and the sticky plant latex make mechanical harvesting difficult. In spite of this, Breemhaar and Bouman (1995) harvested dehiscent genotypes of E. lagascae with a self-propelled pea (Pisum sativum L.) harvester with a customized cutter bar and recovered 70 to 80% of harvestable seeds without incurring machinery problems due to the sticky plant latex, although this required harvesting intact seedpods that were not fully mature.
The objectives of this work were to (i) evaluate the influence of the environment (RH and temperature) on the expression of indehiscence of Vernola grown during different seasons in southern Europe, and (ii) to examine the histological basis of partial dehiscence in capsules from normally indehiscent genotypes.
|
MATERIALS AND METHODS
|
|---|
The indehiscent E. lagascae ‘Vernola’ was used for the trials. Trials were sown in spring and autumn during 2001 and 2002 at two European locations: Murcia in southeast Spain, and Osimo in Central Italy.
Field Experiments in Murcia, Spain
Fields were located at the Torreblanca Experimental Station, 37°47' N and 0°53' W (belonging to Instituto Murciano de Investigación y Desarrollo Agrario y Alimentario), Campo de Cartagena, Murcia, Spain. A 2-by-2-by-2 factorial treatment structure with a split-split plot arrangement, in a randomized complete block design with two replications was set up to study the influence of RH and temperature (during pod formation) on the expression of indehiscence. The factors are explained below.
Irrigation Dose
The average value (1997–2000) of evapotranspiration in an A class tank (ETP0) was used to calculate low (ETP0 x 0.2) or high (ETP0 x 0.8) irrigation doses. In the spring-planted plots, such doses represented 251.4 or 665.2 mm (total rainfall of 24.9 mm during the experiment). Plots established in autumn had ETP values of 197.8 mm (low) and 462 mm (high) and rainfall of 11.2 mm. Water was supplied by a surface drip irrigation system. In Fig. 1
, the schedule of irrigation is shown.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 1. Diagram of irrigation schedule in the field experiments in Murcia, Spain (s = sowing, f = flowering, p = pod formation, and m = maturity).
|
|
Sprinkling
Mister sprinklers were used to increase RH during pod formation. They were located 50 cm above the soil, near the developing capsules within the plants. They operated in cycles of 2 min every 20 min from 0900 until 2000 h. The two levels of this factor corresponded to with sprinkling or without sprinkling.
Plant Density
Low (70 plants m–2) or high (140 plants m–2) densities were tested. Trials were planted on 14 Mar. and 2 Oct. 2001 and harvested on 10 Sept. 2001 and 3 June 2002. This corresponded with a growing season of 180 or 245 d for the spring or autumn vegetative cycles, respectively.
The size of the experimental unit assigned to each treatment was 15 m2 with six rows 5 m long and 0.5 m apart. The layout of the surface drip irrigation system was along the rows with emitters spaced 33 cm apart.
Plots were hand weeded as needed. Plantvax 75 W Educler (Educler, Benifaio, Valencia, Spain; a.i. oxicarboxin, 5,6-Dihydro-2-methyl-1,4-oxathiin-3-carboxanilide 4,4-dioxide) was applied at 2 g L–1 (two treatments at flowering) to control Melampsora euphorbiae (yellow rust). Previcur N [Aventis, Alcacer, Valencia, Spain; a.i. propamocarb, propyl 3-(dimethylamino)propylcarbamate] was applied at 3 mL L–1 (one treatment 45 d after sowing) for Pythium spp. control.
Temperature and percentage RH were monitored for each treatment at 30-min intervals during capsule development by means of sensors recorded by data loggers (testostor 171-3, Testo GmbH and Co., Lenzkirch, Germany) placed in the field at 70 cm over ground level (near plant inflorescences).
Observations and Statistical Analysis
The expression of indehiscence was measured at harvesting, when nearly all pods in a plant were brown. In each plot, 12 plants were sampled (two from each of the six rows). The number of indehiscent capsules per plant were counted and compared with the total number of capsules. Data were obtained for both the spring and autumn trials. Data were analyzed by ANOVAs with GenStat for Windows 5th Edition (VSN International Ltd., UK).
Field Experiments in Osimo, Italy
Test fields were located in Campocavallo Experiment Station, 43°27' N and 13°30' E (belonging to Istituto Sperimentale per le Colture Industriali), at Osimo, in Ancona, Italy. Plots of 10 m2 with densities of 70 plants m–2 (created by thinning) were sown both in spring (30 Apr. 2001) and autumn (23 Oct. 2001); harvested on 5 Oct. 2001 and 24 July 2002, respectively. The resulting length of the growing period amounted to 158 d (spring cycle) or 276 d (autumn cycle).
Rainfall appeared to be enough to supply the water requirements of the crop: 320.2 and 508 mm for the spring and autumn experiments, respectively. Researchers at Oregon State University were able to grow euphorbia without irrigation and <250 mm of natural rainfall in previous experiments (R. Roseberg, 2003, personal communication). No chemical treatments or fertilizers were applied and weeds were removed by hand.
At ripening, 30 plants were sampled. For each plant, the number of indehiscent capsules were counted and expressed as a percentage of the total number of capsules per plant (green capsules, which were in a plant within the 0 to 15.6% range, were discarded). Results were summarized as average capsule retention values and distribution histograms for the plant population sampled.
Capsule Wall Anatomy
In Spain, capsules at different growing stages from indehiscent Vernola plants were sampled. For comparison, capsules of dehiscent wild E. lagascae types grown in an adjacent plot were also sampled.
Harvested capsules were fixed in a 0.1 M cacodilate solution with 3% glutaraldehyde and stored at 4°C for 3 to 5 h. One carpel was detached from each capsule and the seed removed. Carpel walls were rinsed in a cacodilate-sucrose solution and fixed in a 0.1 M cacodilate solution with 1% osmic tetroxide for 1 to 2 h. Afterwards, samples were dehydrated in alcohol series of 50, 70, 90, and 100%. The historesin inclusion technique was applied; samples were infiltrated by soaking them in resin solutions for 2 h at room temperature following an inclusion in harder resin blocks left for polymerization for 2 h at room temperature. Sample orientation transversally or longitudinally was needed to cut the middle sections of the carpel. The blocks with the included samples were sectioned (thickness of 2 µm) with a microtome.
The cuts were stained in toluidine blue solution and mounted for optic microscope observation. Different layers of the capsule wall were identified and their thickness was measured with a micrometer eyepiece.
|
RESULTS AND DISCUSSION
|
|---|
Expression of Indehiscence
There was a significant decrease (P < 0.01) in the number of capsules retained when the plots were sprinkled (Table 1). Indehiscent capsules per plant were reduced from 86.9 to 67.9% in the spring sowing. The interaction between the irrigation dose (I) and sprinking (S) indicates that the described effect was more evident under lower water supply, especially in the high-density plots (Table 1). It seems that water stress on plants (lower doses of irrigation and higher crop densities), together with higher RH (sprinkling), cause partial seed shattering.
The expression of indehiscence in Vernola is not favored by autumn planting in Spain. Overall, 37.9% indehiscent capsules per plant were retained in the autumn cycle against 77.4% in the spring cycle (Tables 1 and 2). Similarly, a very highly significant effect of S (47.2% capsule retention without or 28.5% with sprinkling, respectively) and a significant interaction of I x S (Table 2) were obtained. It should also be noted that autumn-sown plants that were sprinkled were shorter at ripening than those that were not sprinkled, regardless of the other treatments applied.
Sprinkling provided, during the day, conditions of higher RH and lower temperature at ripening (July for the spring-sown plot and April for the autumn-sown plot). For example, at midday in July, 49.7 to 64.5% RH and 26.2 to 30.1°C in comparison with 36.6 to 50.8% and 31.7 to 34.2°C (Table 3). Similarly, in April, at midday values of 44 to 57.8% RH and 23.7 to 25.8°C with sprinkling in comparison with 37.9 to 42.1% and 28.5 to 29.9°C without sprinkling (Table 4). Therefore, we can assume that higher environmental moisture is detrimental for indehiscence expression in Vernola. Holden (1956) indicated that an indehiscent flax variety could partly dehisce 50% of the fruits at a more humid climate. Similarly, Lloyd and Scateni (1968) reported that in grass species, seed shattering is more evident under humid environments.
View this table:
[in this window]
[in a new window]
|
Table 3. Values (at 0.5-h intervals) of relative humidity (RH) and temperature during midday of an average summer day (15–31 July) with or without sprinkling in the spring-sown field plot at Murcia, Spain.
|
|
View this table:
[in this window]
[in a new window]
|
Table 4. Values (at 0.5-h intervals) of relative humidity (RH) and temperature during midday of an average spring day (18–30 April) with or without sprinkling in the autumn-sown field plot at Murcia, Spain.
|
|
It is also possible that moisture had a combined effect on indehiscence with the temperature at the time of pod formation. Comparing both trials without sprinkling (Tables 5 and 6), we observe higher early-morning (0800 h) RH for autumn-sown plots (93.5% in April compared with 69.2% in July) and lower morning temperatures (11.4–29.2°C in April and 19.4–33.9°C in July).
View this table:
[in this window]
[in a new window]
|
Table 5. Values (at 4-h intervals) of relative humidity (RH) and temperature of an average summer day (15–31 July) with or without sprinkling in the spring-sown field plot at Murcia, Spain.
|
|
View this table:
[in this window]
[in a new window]
|
Table 6. Values (at 4-h intervals) of relative humidity (RH) and temperature of an average spring day (18–30 April) with or without sprinkling in the autumn-sown field plot at Murcia, Spain.
|
|
Conditions of greater environmental moisture, lower temperature, or both combined, either during seed pod formation or ripening, may cause a poorer indehiscence expression in Vernola, as it has been observed especially in the sprinkled plots of the autumn-sown fields. Singh and Bejiga (1991) associated pod dehiscence in chickpea with more number of days to maturity. Pascual-Villalobos et al. (1994) suggested that shorter vegetative periods favor expression of histological characteristics of E. lagascae capsules responsible for indehiscence. Young (1986) reported that seed shattering in small buffalo grass (Panicum coloratum L.) was more severe in wet environments than in drier climates at the time of flowering. Stepchenko (1986) indicated that weather conditions, date of harvest, and the cultivar influence pod shedding in flax. Altogether, the literature examples and our own observations suggest that longer crop periods and higher moisture and lower temperature conditions during flowering and pod formation favor dehiscence.
At Osimo, less capsules were retained per plant in the autumn- (67.2%) than in the spring-sown (78.8%) trial (Table 7). Comparing both locations in Spain (without sprinkling) and Italy (Tables 1, 2, and 7), greater indehiscence was obtained in Spain (86.9 and 78.8%) for the spring-sown trial but reversed (47.2% and 67.2%) for the autumn trials. Comparing the climatic conditions (Tables 5, 6, and 8), we see that temperature during capsule development was lower in Italy; however, this single factor cannot explain the results by its own and it seems that other factors might as well have an influence.
View this table:
[in this window]
[in a new window]
|
Table 8. Temperature and relative humidity (RH) in Osimo, Italy, during the spring (May–September 2001) and autumn (October 2001–July 2002) growing seasons.
|
|
Figure 2
shows distribution histograms of plants grown at Osimo location. If plants are sown in spring, they fall within the 70 to 90% interval of retained capsules, while plants sown in autumn were more frequently found in the 50 to 70% indehiscence range.
Table 8 is a summary of the climatic conditions during the field experiments in Italy. It is interesting that from November to April, plants were under temperatures below zero (up to –9.3°C) and some of them survived, which can be used for breeding for frost tolerance.
Roseberg (2000) reported unpredictable shattering effects from late summer and fall plantings in Oregon, USA, although this observation changed from year to year during his study period of 1997 to 1999. Higher doses of irrigation and wider row spacings produced more shattering in Vernola. Growth was reduced at 10°C but in spite of this, plants could survive low daily winter temperatures around –12°C, although they died at –17°C in Oregon.
Capsule Wall Anatomy
In Fig. 3
, the histology of transversal sections of carpel walls in indehiscent capsules is displayed. The young pericarp is thicker (325 µm) at the point of union (a) between carpels in comparison with the kill (b) (225 µm) in which a suture is found (Fig. 3A). When ripening occurs, cracks (c) appear in this latter area (see Fig. 3B). Meakin and Roberts (1990) pointed out that in oilseed rape (Brassica napus L.) siliqua, an abcission layer at valve edges is responsible for cell wall degradation and dehiscence.
In dehiscent capsules of E. lagascae wild types, a middle layer (mesocarp) with elongated cells (d) orthogonal to outer and inner pericarp layers is present all over the wall (Fig. 4A)
. However, only a small portion of the mesocarp remains (e) just at carpel kill in transversal sections of indehiscent Vernola capsules (Fig. 4B). Our interpretation is that the fully developed mesocarp is responsible for a strong shattering in capsules of dehiscent genotypes, while in indehiscent genotypes this layer remains undeveloped near the sutures in the capsules, making them susceptible to a slower opening under some environmental circumstances. Ohta et al. (1991) described that the expansion of hypodermal cell layers in humid conditions causes splitting of the cuticule in many fleshy fruits. Piccirilli and Falcinelli (1989) studied the histological characteristics in orchardgrass (Dactylis glomerata L.) and they observed two abcission layers, one in the rachilla and the other in the pedicel in which the mechanical and enzymatic deterioration process occurs with differences between high and low shattering varieties.
In Fig. 5
, a plate of a longitudinal cut shows that the pericarp is thicker in the upper (f) part (337 µm) of the capsule and becomes thinner (200 µm) toward the pedicel (g). Therefore, the easiest direction for opening is bottom up along the carpel as it really occurs. In castor (Ricinus communis L.), the shells are also thicker in the upper part of the capsules (Moshkin, 1986).
The comparison of sections (Fig. 6)
reveals that the mesocarp in dehiscent genotypes (h) is wider (206 ± 6 µm) than in indehiscent (i) Vernola (88 ± 7.2 µm). In Pistacia atlantica Desf., a reduced shell splitting results from thinner shells in comparison with the dehiscent P. vera L. (Shuraki and Sedgley, 1996). The mesocarp oriented at a 90° angle to the exocarp cells provokes tensions at the time of drying and subsequent pod opening in Lotus (Young et al., 1990). We propose that in euphorbia, this tension results in violent shattering and seed expulsion observed in dehiscent genotypes while in indehiscent Vernola, weaker tensions are also able to cause cracks (Fig. 4B and Fig. 3B) and eventually produce some partial dehiscence (see data of Tables 1, 2, and 7) in the indehiscent variety.
In previous studies (Pascual-Villalobos et al., 1994), the absence of the mesocarp and a reduced capsule wall weight in indehiscent genotypes was reported. It can be concluded that a reduced mesocarp is present in Vernola and it is the physical reason of the partial dehiscence observed in our trials.
From the spring sowing in Spain, heavier capsule walls (t test = 5.995, P < 0.05) were obtained for plants grown with sprinkling (dry weight for five capsules = 0.1072 g) in comparison without sprinkling (0.0828 g); such result was not obtained when plants were sown in autumn. However, no histological differences were observed under the microscope between capsules of both treatments.
Photographs of Vernola capsules (Fig. 7)
show the splitting zone from the bottom and along the dorsal suture in the carpel. The process in time goes from (j) to (k) and finally (l). Such partial dehiscence is explained by the histological features described in Fig. 3 to 6. According to Peng and Tobe (1987), in Ludwigia (Onagraceae), a loculicidal dehiscence occurs when the opening is along the midrib of each carpel.
Grant (1996) pointed out that the internal stress of fruits at drying is responsible for dehiscence. Greater moisture in thicker layers will produce greater humidity loss and tissue shrinkage and therefore more stress for shattering. Winch and MacDonald (1961) also stated that pod dehiscence is caused by differential rates of moisture loss in tissues.
Thus, it appears that the Vernola cultivar of E. lagascae has the potential to be grown so that dehiscence is minimized. However, cool and moist conditions during flowering, pod formation, and ripening may increase the size of the mesocarp and capsule wall, thus resulting in a morphology more similar to the dehiscent wild types, and a greater percentage of dehiscent pods in the field.
|
ACKNOWLEDGMENTS
|
|---|
This work has been funded by the V Framework Programme of the European Union: Stays at Marie Curie Training Sites (Contract no. QLK5-1999-00355). Francesco Verdolini (March 2001–February 2002) and Andrea Anconetani (March 2002–February 2003) were awarded one year research fellowships at Instituto Murciano de Investigación y Desarrollo Agrario y Alimentario. We thank R. Roseberg, from Oregon State University, Southern Oregon Research and Extension Center Research Unit, for critical review of the manuscript.
Received for publication May 29, 2003.
|
REFERENCES
|
|---|
- Breemhaar, H.G., and A. Bouman. 1995. Harvesting and cleaning Euphorbia lagascae, a new arable oilseed crop for industrial application. Ind. Crops Prod. 4:173–178.
- Cromack, T. 1999. The performance of Euphorbia lagascae in southern England. p. 31–37. In The development of Euphorbia lagascae as a new oil crop within the European Community (FAIR CT98/4460). Proc. of Workshop I, Sept. 1999. CPL Scientific Publishing Services, UK.
- Fahn, A., and M. Zohary. 1955. On the pericarpial structure of the legumes, its evolution and relation to dehiscence. Phytomorphology 5:99–111.
- Grant, W.F. 1996. Seed pod shattering in the genus Lotus (Fabaceae). A synthesis of diverse evidence. Can. J. Plant Sci. 76(3):447–456.
- Holden, D.J. 1956. Factors in dehiscence of the flax fruit. Bot. Gaz. (Chicago) 117:294–309.
- Lloyd, D.L., and W. Scateni. 1968. Makarikari grasses for heavy soils. Queensl. Agric. J. 94:721–724.
- Meakin, P.J., and J.A. Roberts. 1990. Dehiscence of fruit in oilseed rape (Brassica napus L.). I. Anatomy of pod dehiscence. J. Exp. Bot. 41(229):995–1002.[Abstract/Free Full Text]
- Moshkin, V.A. 1986. Castor. Amerind Publ. Co., New Delhi, India.
- Ohta, K., N. Ito, T. Hosoki, and Y. Suyi. 1991. Influence of relative humidity on fruit cracking of cherry tomato grown in hydroponic culture and its control. J. Am. Soc. Hortic. Sci. 60:337–343.
- Pascual-Villalobos, M.J. 1999. Registration of indehiscent Euphorbia lagascae L. germplasm: Idm24, idm76, and idm77. Crop Sci. 39:1537.
- Pascual-Villalobos, M.J., G. Röbbelen, and E. Correal. 1994. Production and evaluation of indehiscent mutant genotypes in Euphorbia lagascae. Ind. Crops Prod. 3:129–143.
- Peng, C., and H. Tobe. 1987. Capsule wall anatomy in relation to capsule dehiscence in Ludwigia sect. microcarpum (Onagraceae). Am. J. Bot. 74(7):1102–1110.
- Perez Marco, P. 1999. Performance and research needs for Euphorbia lagascae crop under mediterranean conditions. p. 47–52. In The development of Euphorbia lagascae as a new oil crop within the European Community (FAIR CT98/4460). Proc. of Workshop I, September 1999. CPL Scientific Publishing Services, UK.
- Piccirilli, M., and M. Falcinelli. 1989. Anatomy of seed dispersal mechanisms in high and low seed shattering cultivars of orchardgrass. Crop Sci. 29:972–976.[Abstract/Free Full Text]
- Roseberg, R. 2000. Agronomic research on Euphorbia lagascae in Oregon, USA. p. 58–75. In The development of Euphorbia lagascae as a new oil crop within the European Community (FAIR CT98/4460). Proceedings of Workshop II, March 2000. CPL Scientific Publishing Services, UK.
- Shuraki, Y.D., and M. Sedgley. 1996. Shell structure and embryo development of Pistacia vera L. and P. atlantica Desf. (Anacardiaceae) following intra- and interspecific pollination. Int. J. Plant Sci. 157(5):586–594.
- Singh, K.B., and G. Bejiga. 1991. Evaluation of world collection of chickpea for resistance to pod dehiscence. J. Genet. Breed. 45:93–96.
- Stepchenko, A.G. 1986. The resistance of forms to seed shattering. Len Konoplya 2:37–38.
- Turley, D., M. Froment, and S. Cook. 2000. Development of Euphorbia lagascae as a new industrial oil crop. ADAS, Wolverhampton, UK.
- Winch, J.E., and H.A. MacDonald. 1961. Flower, pod and seed development relative to the timing of the seed harvest of the viking birdsfoot trefoil (Lotus corniculatus L.). Can. J. Plant Sci. 41:523–532.
- Young, B.A. 1986. A source of resistance to seed shattering in Kleingrass (Panicum coloratum L.). Euphytica 35:687–694.
- Young, J.B., D.A. Somers, R.L. Wright, and R.L. McGraw. 1990. Seed pod dehiscence in birdsfoot trefoil, Lotus conimbricensis, and their interspecific somatic hybrid. Can. J. Plant Sci. 70:279–284.
TOP
ABSTRACT
INTRODUCTION
