Plant Cell, Tissue and Organ Culture 36: 227-236, 1994. (~) 1994 Kluwer Academic Publishers. Printed in the Netherlands.
Anther and microspore culture of Lupinus albus in liquid culture medium A. J. O r m e r o d & P. D. S. Caligari*
Department of Agricultural Botany, School of Plant Sciences, University of Reading, Whiteknights, PO Box 221, Reading, Berks, RG6 2AS, UK (*requestsfor offprints) Received 31 August 1992; accepted in revised form 17 October 1993
Key words: cytology, embryo-like-structures, lupins, pollen
Abstract
Embryo-like structures (ELS) with clearly defined cotyledons and radicles have been regenerated from
Lupinus albus L. microspores. ELS induction from microspores in liquid medium was successful, with a maximum of over 3,500 ELS per anther being produced from microspores predominantly at the early binucleate stage of development. Cytological analysis revealed that first pollen grain mitosis occurred in closed buds with maximum ELS production being obtained from buds at the point of first petal emergence. Generally there was a lack of synchronisation of microspores within anthers from the less mature bud development stages. Mechanical isolation and culture of microspores proved to be unsuccessful. Direct germination from ELS has not been observed; recurrent somatic embryogenesis occurred following internal cleavage within the ELS or on the surface of the ELS. Methods of increasing the level of mature ELS capable of germination are under investigation.
Abbreviations: BA-6-benzyladenine, 24-D-2,4-dichlorophenoxyaceticacid, NAA-naphthaleneacetic acid, 2iP - N-isopentyl-adenine, GA3 - gibberellic acid, S&H - Schenk & Hildebrandt medium (1972), M&S - Murashige & Skoog medium (1962); N&N - Nitsch & Nitsch medium (1969), B5 - Gamborg's B5 medium (1968), NNB5 - N&N salts plus B5 vitamins, ELS - Embryo Like Structures, DAPI 4',6-Diamidino-2-Phenylindole Introduction
There is continuing interest in developing a range of grain lupin species because, with the exception of soybeans, they contain higher levels of seed protein than the other grain legume species (Williams 1984). Lupinus albus L. seeds contain between 34 45% protein and in addition 10-15% oil while, at the same time, being well adapted to low input systems and impoverished soils covering a range of pH values. In addition to their ability to fix atmospheric nitrogen there is interest in their efficiency, compared to other crop plants,
in making available immobile phosphates to following crops, particularly in soils containing low levels of available phosphates. The relatively poor rate of progress in legume biotechnology and tissue culture has been ascribed to a low level of investment in research (with the exception of soybeans) and the lack of reproducible, fast and easy to handle protocols for plant regeneration (Jacobsen 1992). Anther and microspore culture techniques would offer rapid methods for recovering homozygous inbred lines and thus accelerate the production of new cultivars. Attempts have been made to develop anther
228 culture systems for a range of lupin species (Sator et al. 1983; Sator 1985, 1986; Campos-Andrana & Mota 1989). Plants have only been successfully reported to have been regenerated from two anther-derived calli from L. polyphyllus (Sator 1985). Identification of the optimum microspore developmental stage for anther and microspore culture is an important prerequisite of the culture process. Small differences in developmental age produce great differences in yield (Dunwell 1986). The present paper reports the staging of microspore development within buds and the production of L. albus ELS following spontaneous microspore release from anthers in liquid medium.
large and five small anthers; the length of the large anthers was recorded. A single large anther from each flower bud was crushed in a drop of DAPI prepared using the method of Detchepare et al. (1989). The same treatment was applied to the five small anthers, which were pooled because individually they contained fewer microspores than the large anthers. After debris was removed, slides were covered, sealed and stored in the dark at 4°C, to allow the uptake of dye, and then observed. Observations were made under a Leitz Dialux fluorescent microscope. Microspore development was observed from flower buds derived from main stem and primary branches, harvested in August and February. Induction of ELS in liquid medium
Materials and methods
Plant material L. albus cultivars and lines representing a wide range of types were used in the experiments. Accessions came from the Unviersity of Reading collection with the addition of seven lines supplied by Dr. Christian Huyghe, SAPF-INRA, Lusignan, France. Seeds were imbibed for 24 h and coated with a dressing containing Bradyrhizobium lupinii, prior to vernalization for 3 weeks at 10°C in plastic bags containing damp perlite. The seedlings were transferred to 18 cm pots containing a 50:50 mixture of John Innes No. 1 compost and perlite. The planted pots were placed in the glasshouse, under a 16-h photoperiod provided by high pressure sodium lighting producing an average of 370-450 ~tmol m -2 s- 1 of light across the bench. Plants were grown throughout the year under glasshouse conditions while field plots were only grown during the summer months.
Flower buds were washed in distilled water prior to surface disinfestation with a 5% sodium hypochlorite solution for 10 min. Buds were rinsed three times with sterile nanopure water. Buds were graded and stages A to H (Fig. 1) were placed in separate Sterilin 90 mm petridishes containing 2 ml of the culture medium. The medium used (termed NNB5) consisted of Nitsch & Nitsch (1969) macroelements, Gamborg B5 microelements (Gamborg et al. 1968), supplemented with 6.95x10-aM L-proline, 9.51× 10-4M L-serine, 5.37x 10-6M NAA, 4.65x 10-6M kinetin, 2.22x 10-6M BA, 2.26 x 10-6M 2,4-D and either 5% sucrose or maltose. The pH of the medium was adjusted to 5.8 and autoclaved at 121°C (103.4 kPa) for 20 min. Stamens from each flower bud were cultured with or without carpels individually in 30 mm Nunc petri-dishes containing 0.5 ml of medium in the dark at 25°C. ELS developing from microspores in culture were stained with DAPI.
Staging bud and microspore development
Attempts to promote ELS maturation on solid and in liquid media
Flower buds were graded according to size and development (Table 1 and Fig. 1). Overall bud length and length of visible petal were recorded prior to dissection. L. albus flowers contain five
Developing ELS were transferred to 90 mm Sterilin petri-dishes containing S&H medium containing 5% sucrose or maltose and solidified with 0.8% Bacto-agar (Difco), supplemented with con-
229 Table 1. Size of L. albus and anthers (data from seven accessions, February 1992). Character measured
A
B
C 7 7.4 -4- 0.32
Bud length (mm)
n :~z
1 2.2
7 5.3 -4- 0.45
Petal length (mm)
n Y:
.u
_ -
Main anther length (mm)
n :~
Colour
<1.0 green
Stage of development D E-F
G
H
9 8.4 4- 0.45
7 12 4- 0.32
8 15.1 4- 0.63
7 15.5 -4- 0.55
9 4.7 4- 0.45
7 8.6 4- 0.45
8 11.6 4- 0.71
7 11.7 ± 0.89
35 1.3 + 0.1
30 1.9 -4- 0.1
45 2.1 4- 0.1
30 2.3 ± 0.1
40 1.9 q- 0.1
green/ yellow
yellow
yellow
yellow
yellow
z = Mean (mm) -4- standard error. u = All buds closed in stages A - C
Fig. 1. L.albus bud and anther development stages A to H.
centrations of NAA between 5.37x 10-6M and 5.37x 10-SM and of BA between 4.44x 10-6M and 4.44x 10-SM. The plates were cultured at 25°C. After 8 weeks, ELS were transferred from S&H medium to NNB5 liquid or solid culture medium containing 5% maltose with or without the growth regulators used for microspore induc-
tion. In addition, the response of ELS was tested in NNB5 medium supplemented with 10 - 4 , 10 -5, 10-6M NAA, kinetin, BA, 2iP, zeatin or 2.63 × 10-TM GA3 with or without agitation (on a rotary shaker at 80 rpm) and with or without the addition of 100~tl per dish of a warm solution consisting of 1.0 g 1-1 activated charcoal and 0.5 g 1-1 agar solution (Table 2). There
230
Table 2. Number of plates (out of 5) per treatment containing ELS resembling those found in white lupin seeds in vivo, following transfer from S&H solid medium to NNB5 liquid medium supplemented with a range of growth regulators and charcoal with (+) or without (-) agitation. Agitation charcoal NAA + + +
+ Growth regulator type conc(M) × 10 -6 "t × 10 -6 ~ 0
NAA Kin BA 2,4-D
5.4 4.6 2.2 2.3
Ga3
2.6 × 10 -7
× 10 - 6
x 10 -6 J 3
None
0
-
+
1
0
1
2 0
1 1
2 1
+
+
10 -5
2x10 -5
%
+ 10 -5
10 -5
10 -4
-~
10
40 10
Zea Zea
10 -6 10 -5
0 0
0
0
2
20
Zea
10 -4
0
0
0
Kin
10 -6
0
Kin Kin
10 -5 10 -4
1 0
0 0
1 0
0 0
1 1
0 0
7 7
1
1
0
0
1
10
BA
10 -6
0
3
0
0
0
0
BA BA
10 -5 10 -4
0 0
0
1
0
0
0
-
3
1
0
2
0
0
-
10
2iP 2iP
10 -6 10 -5
0 1
1
-
-
10
2
-
-
30
2iP
10 -4
0
2
-
-
20
Kin Kin Kin
2 x 10 -5 10 -5 10 -4
None %~o
6.7
16.0
11.0
13.0
6.7
3.3
10
1v
1
1
15
3 2
2 2x
1 1
40 38
2
0
2
27
40
28
20
x treatment combination not tested u based on 10 plates per replicate based on 3 plates per replicate ~o % values based only on treatment combinations tested.
were and
five cultures all w e r e
4 weeks. tures,
present
grown retained
treatment
i n t h e l i g h t at 2 5 ° C
Subsequently,
were
for each
for
three of the five cul-
at 2 5 ° C
for a further 4
weeks
w h i l e the o t h e r t w o w e r e t r a n s f e r r e d to
4°C. E L S f r o m e a c h t r e a t m e n t , as i n d i c a t e d in Table 2, w e r e t r a n s f e r r e d t o e i t h e r M & S , S & H , N & N o r
231 B5 media containing 3% maltose, pH 5.8 and solidified with 0.8% agar and cultured at 25oC under a 16-h photoperiod (34-50 ~tmol m -2 s -1) provided by diffuse cool white fluorescent tubes.
Statistical analysis The effects of bud development stage on ELS production were analysed using Chi-squared, while the effects of the different treatments designed to promote embryo maturation in NN69B5 liquid medium were analyzed using the General Linear Model and Categorical procedures in SAS (Statistical Analysis System) software.
Results
Staging bud and microspore development Buds and anthers within a range of developmental stages proved to be uniform in size (Table 1). First pollen grain mitosis was observed to occur prior to flower bud opening. Microspores with dividing nuclei were observed at bud stages B-C in buds harvested in August; binucleate microspores were observed at bud stage B in buds harvested in February. The point of first pollen grain mitosis appeared not to be completely synchronous in the buds sampled. At and just after bud opening, it appeared that most microspores were at the early to mid-binucleate stage. Bud stages E-G contained microspores at the mid- to late-binucleate stages. Spindle-shaped nuclei were observed in microspores from the beginning of bud opening and they were predominant at bud development stages E-G.
Induction of ELS in liquid medium No response was obtained from anther or microspore cultures initiated during the spring or summer months during which period anthers were observed to release microspores into the medium, but no ELS developed from them. ELS were produced on all four occasions during the autumn and winter months from anthers cultured
in liquid medium. Microspores from anthers taken from bud development stages E and G exhibited several types of changes in culture medium. The types of changes observed included: enlarged cells with exine and intine separating from the cytoplasmic contents; abnormal pollen tube formation; or protrusion of cytoplasmic contents from the colpi. Some microspores formed clumps and it also appeared that some of the microspores had undergone division to produce groups of large cells. Microspores from bud stages B to E released embryo-like structures into the liquid medium. Staining with DAPI revealed that they were composed of tiny cells, characteristic of embryogenic development. It appears that rapid cell division without cellular expansion had occurred before the ELS were released from the microspores via one of the three colpi. Microscopic examination indicated that many of the L. albus microspore-derived structures possessed polarity ab initio. Globular (Fig. 2C), rod/torpedo (Fig. 2E, and heart-shaped ELS were observed developing in liquid medium, development was not synchronous and new ELS were produced over several weeks. In addition, organised bipolar structures, consisting of large cells, were produced from anthers at stage F (Fig. 2F). These structures did not, however, show any further subsequent development. Chi-squared analysis showed that there were no significant differences in ELS production between the range of bud development stages, possibly reflecting the small sample size analysed. The overall trend suggested that highest percentage of plates producing ELS were obtained by culturing buds around the point of first bud opening (Stages C/D, D and D/E), as seen in Fig. 3. In one culture from stage D buds, spherical masses of cells were observed breaking out of the microspore wall after only 2 days in culture (Fig. 2B). These continued to develop in size (Fig. 2D). On the first occasion when microspore callus was obtained, a total 273 ELS (equivalent to eight ELS per anther) were obtained in one of the dishes at stage D (Fig. 2A). In a later trial 7,800 ELS per ml were produced from one stage C plate, being equivalent to 3,546 ELS per anther. Sucrose and maltose were found
232
Fig. 2.
233
Fig. 2. In vitro development of L. albus microspore-derived ELS. (A) ELS developing from microspores released from cultured anthers. (B) ELS composed of fine cells escapingfromexine. (C) GlobularELS with suspensor like structure attached. (D) ELS increasing in size in liquid medium. (E) Rod/torpedoshaped ELS. (F) Structure formed of large cells, derived from bud stage F microspores. (G) Somaticembryos being produced on surfaces of ELS. (H) Recurrent somatic embryogenesis following internalcleavage in the ELS. (I) Heart shaped ELS. (J) ELS with developingradicle resemblingtypes foundin vivo.
to be equally effective, since sucrose produced slightly fewer ELS than maltose but they were larger. Attempts to promote ELS maturation on solid and in liquid media
A total of 273 ELS (from above) were transferred to 10 plates of S&H medium in an effort to promote further ELS growth and differentiation. The ELS grew to form cream-coloured masses consisting of small cells, and underwent recurrent somatic embryogenesis following cleavage within the ELS. In addition, somatic embryogenesis occurred with small cell masses sloughing off from these larger ELS and growing to form globular ELS. After 8 weeks, over 2,000 distinct ELS had been formed from the original 273. Anthers were also transferred from the liquid medium to the solid S&H medium and creamy sheets of fine cells were produced that covered the anthers. These sheets of cells were easily separated from the anther wall and so seemed not to be attached to, or formed from, the anther walls (which
themselves showed no signs of callus formation), but probably originated from microspore callus released from within the anther via the stomia. Transfer of ELS to NNB5 solid medium produced similar results to those obtained on S&H solid medium. Bi-polar ELS up to 2 mm were observed in all the liquid medium, with a range including globular, heart and torpedo shaped (Fig. 2). New globular ELS were initiated from cells sloughing off existing ELS. ELS were observed either to divide in culture (Fig. 2H), or less commonly form new globular ELS on their surfaces (Fig. 2G). Some ELS were observed to have passed the torpedo stage and were a similar shape to zygotic embryos 28-34 days post anthesis (Fig. 2J). There were significantly more plates containing these more mature ELS when 2.63 x 10-7M GA3 was incorporated in the medium, compared to all other growth regulator treatments. However, the numbers of these ELS per plate were low and even in the GA3 treatment amounted to no more than three structures per plate. Agitation and addition of charcoal had no significant effect on ELS maturation (Table 2).
234 70
50
ra~ ['~
40
,N
~.
3o
20
10
0
I
A
B
B/C
C
C/D Bud
D
DIE
E
F
G
Stage
Fig. 3. P e r c e n t a g e s o f petri dishes p r o d u c i n g ELS f r o m different stages o f bud d e v e l o p m e n t .
No significant differences were observed in development of ELS following transfer to the different solid basal medium. Development continued as in the liquid medium, and many of the more advanced ELS also divided. A low level of ELS developed further, without splitting; cotyledons continued to expand and the radicles were well defined. However, further development and germination has so far not occurred.
Discussion
First pollen grain mitosis has so far been observed to occur as early as bud stages A-B and as late as stage C just before bud opening. As in L. polyphyllus (Sator et al. 1983), there was generally a lack of synchronisation among the microspores from a single anther, particularly in the less mature stages
of bud development. Further analysis will be needed to determine if the pattern of microspore development varies through the season. Previous work has shown only limited success in ELS production in lupins (Sator et al. 1983; Sator 1985, 1986; Campos-Andrana & Mota 1989). In the present study it has been possible to produce ELS following spontaneous microspore release from L. albus anthers initially cultured in NNB5 liquid medium. Spontaneous release of microspores from anthers at a bud development stage conducive to ELS production is not possible from all species, for example, Brassica napus (Lichter 1981). L. albus anthers release their microspores into liquid medium in a similar fashion to Nicotiana tabacum (Sunderland & Roberts 1977). Anthers from bud stages A-C released very few microspores into the liquid culture medium. The highest number of ELS
235 was produced from anthers containing mainly young binucleate microspores around the point of first petal appearance (stage D) as shown in Fig. 3. In one culture at this stage of development, microspore-derived ELS were observed after only 2 days of incubation. These results may suggest that the time for microspore release in culture decreases with anther age as has been reported in tobacco (Sunderland & Roberts 1977). Large quantities of microspores were released from older anthers from bud stages E-G, but did not develop into ELS. Mechanical isolation and culture of microspores has so far proved largely unsuccessful. Similarly, direct mechanical isolation and culture in sugar containing media has not been reported without preculturing anthers or other pretreatments (Wernicke & Kohlenbach 1977; Wernicke et al. 1978; Kyo & Harada 1985, 1986). Microspore-derived structures found in L. albus were generally embryogenic in origin consisting of tiny cells; similarly ELS produced by Nicotiana tabacum were reported to be comprised of small compact cells, due to the rate of cell division being more rapid than cell enlargement (Sunderland & Wicks 1971; Imamura & Harada 1980). A few ELS have been produced with defined radicles and cotyledons that resembled those found in vivo during the rapid phase of embryo development. Somatic embryos recurrently proliferated from the ELS; a similar situation occurs in Brassica napus where less than 5% of ELS have been reported to produce shoots directly from primary embryo shoot meristems (Swanson et al. 1987). So far our efforts to find the key to switching from recurrent somatic embryogenesis to maturation and germination have been unsuccessful. Inclusion of 10-4M kinetin alone or in combination with low concentrations of IAA was reported to be an effective method of inducing leaf initiation (Loh et al. 1983), but none of the cytokinin or auxin treatments so far tested have prevented recurrent somatic embryogenesis from ELS. GA3 used to promote shoot production from Brassica napus ELS (Chuong et al. 1988; Kott et al. 1988) was also unsuccess-
ful with L. albus ELS. Combinations of ABA and cytokinins have been used with and without GA3 to promote embryo maturation and germination by Ammirato (1977) and Bozhkov et al. (1992). A similar approach may be applicable to obtain maturation and germination of L. albus ELS. The range of ELS types observed resembel those reported from Brassica spp. microspore culture (Chuong & Beversdorf 1985). The mineral salt composition of the media may also influence ELS development and has been reported to affect cotyledon development in zygotic embryos (Monnier 1990). It may be that such alterations in the culture conditions will increase the level of production of desirable ELS types. ELS have developed so far only from anthers cultured during the autumn and winter months. Further investigations are being conducted to determine whether environmental factors, such as duration and quality of illumination and temperature, affect the donor plants. The most reproducible results from Nicotiana rustica microspore culture were obtained from plants grown mainly in the winter (November to February) (Kyo & Harada 1985). Similarly, short day lengths and low temperatures have been reported to increased pollen plant production from some genotypes of Nicotiana tabacum (Heberle-Bors 1984). These initial results demonstrate that it is possible to regenerate defined ELS from L. albus microspores and the rate of production of ELS from the optimum stage of microspore development is promising for the future use of this technique as a tool in breeding new white lupin varieties.
Acknowledgements This work has been made possible by financial support from the Ministry of Agriculture, Fisheries and Food. We are also grateful to thank Dr. Christian Huyghe, INRA, Lusignan for providing seven L. albus accessions and Dr. John Day, IACR, Rothamsted for the Bradyrhizobium lupinii dressing.
236 References Ammirato PV (1977) Hormonal control of somatic embryo development from cultured cells of caraway. - interactions of abscisic acid, zeatin and gibberellic acid. Plant Physiol. 59:579-8 Bozhkov PV, Lebedenko LA & Shiryaeva (1992) A pronounced synergistic effect of abscisic acid and 6benzyladenine on Norway spruce (Picea abies L. Karst) somatic embryo maturation. Plant Cell Rep. 11:386-389 Campos-Andrana MP & Mota M (1989) Influencia do meio de cultura na inducau de callus e na diferenciacao de raizes e caules a partir da cultura de anteras de Lupinus hartwegii Lindl. XXIV Joznadas de Genetica Luso-Espanolas, Evora (Portugal) Oct 1989 (p 19) Chuong PV & Beversdorf WD (1985) High frequency embryogenesis through isolated microspore culture in Brassica napus L. and B. carinata Braun. Plant Sci. 39: 219-226 Chuong PV, Pauls KP & Beversdorf WD (1988) High frequency embryogenesis in male sterile plants of Brassica napus through microspore culture. Can J. Bot. 66: 16761680 Detchepare S, Heizmann P & Dumas C (1989) Changes in protein patterns and protein synthesis during anther development in Brassica oleracea. J. Plant Physiol. 135:129137 Dunwell JM (1986) Pollen, ovule and embryo culture, as tools in plant breeding. In: Withers LA & Alderson PG (Eds) Plant Tissue Culture and its Agricultural Applications (pp 375-404). Butterworths, London Gamborg OL, Miller RA & Ojima K (1968) Nutrient requirement of suspension cultures of soybean root cells. Exp. Cell Res. 50:151-158 Heberle-Bors E (1984) Genotypic control of pollen plant formation in Nicotiana tabacum L. Theor. Appl. Genet. 68:475-479 Imamura J & Harada H (1980) Effects of abscisic acid and water stress on the embryo and plantlet formation in anther culture of Nicotiana tabacum cv. Samsun. Z. Pflanzenphysiol. 100:285-289 Jacobsen HJ (1992) B iotechnology applied to grain legumes current state and prospects. Proceedings of the 1st European Conference on Grain Legumes, Angers, France (pp 99-103). European Association for Grain Legume Research, Paris Kott LS, Polsoni L & Beversdorf WI (1988) Cytological aspects of isolated microspore culture of Brassica napus. Can. J. Bot. 66:1658-1664 -
Kyo M & Harada H (1985) Studies on conditions for cell division and embryogenesis in isolated pollen culture of Nicotiana rustica. Plant Physiol. 79:90-94 Kyo M & Harada H (1986) Control of the developmental pathway of tobacco in vitro. Planta 168:427--432 Lichter R (1981) Anther culture of Brassica napus in a liquid culture medium. Z. Pflanzenphysiol 103:229-237 Loh C-S, Ingram DS & Hanke DE (1983) Cytokinins and the regeneration of plantlets from secondary embryoids of winter oilseed rape, Brassica napus ssp oleifera. New Phytol. 95:349-358 Monnier M (1990) Zygotic embryo culture. In: Bhojwani SS (Ed) Development in Crop Science 19: Plant Tissue Culture: Applications and Limitations (pp 366-393). Elsevier, Amsterdam Murashige T & Skoog F (1962) A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol Plant. 15:473-497 Nitsch JP & Nitsch C (1969) Haploid plants from pollen grains. Science 163:85-87 Sator C, Mix G & Menge U (1983) Investigation on anther culture of Lupinuspolyphyllus. Plant Res. Dev. 18:37-46 Sator C (1985) Regeneration von Lupinpflanzen aus antheren. Landbauforschung Volkenrode 35:5-7 Sator C (1986) Chances of crop improvement in lupins using tissue culture techniques. Report of the IVth International Lupin Conference (pp 180-189). Geraldton, Australia Schenk RU & Hildebrandt AC (1972) Medium and techniques for inductionand growth of monocotyledonousand dicotyledonousplant cell cultures. Can J. Bot. 50:199-204 Sunderland N & Roberts M (1977) New approach to pollen culture. Nature 270:236-238 Sunderland N & Wicks F (1971) Embryoid formation in pollen grain ofNicotiana tabacum. J. Exp. Bot. 22: 213226 Swanson EB, Coumans MP, Wu SC, Barsby TL & Beversdoff WD (1987) Efficient isolation of microspores and the production of microspore derived embryos from Brassica napus. Plant Cell Rep. 6:94-97 Wernicke W & Kohlenbach HW (1977) Versuche zur kultur isolierter mikrosporen von Nicotiana and Hyoscyamus. Z. Pflanzenphysiol. 81:330-340 Wernicke W, Harms CT, Lorz H & Thomas E (1978) Selective enrichment of embryogenic microspore populations. Naturwissenschaften 65:540-541 Williams W (1984) Lupins in crop production. Outlook on Agriculture 13:69-76
