PHENOTYPIC WEST (MusA

DivEmsrrY

AND CENTRAL SPP., AAB

AND PATTERNS

AFRICAN

GROUP

RONY SWENNEN,D I R K

OF VARIATION

IN

PLANTAINS

MUSACEAE)

1

VUYLSTEKE, AND R O D O M I R O O R T I Z

Swennen, Rony (Katholieke Universiteit Leuven, Laboratory of Tropical Crop Husbandry, Kardinaal Mercierlaan 92, 3001 Heverlee, Belgium), Dirk Vuylsteke (East & Southern African Regional Center, International Institute of TropicalAgriculture (IITA), P.O. Box 7878, Kampala, Uganda), and RodomiroOrtiz (Corresponding author. International mailing address: HTA, c/o L. W. Lambourn & Co., Carolyn House, 26 Dingwall Road, Croydon, CR9 3EE, England. Plantain & Banana Improvement Program, HTA High Rainfall Station, PMB 008, Nchia Eleme, Rivers State, Nigeria). PHENOTYPXCDIVERSITYAND PAvrP-mNSOF VARIATIONIN WESTAND CENTRAL AFRICAN PLANTAINS(Mvsa see., AAB GROUPMUSACEAE).Economic Botany 49(3): 320--327. 1995. Plantains (Musa spp., AAB group) are an important food crop and an integral component of the farming systems in the lowland humid forest zone of West and Central Africa. A group of 24 plantain cultivars, representing the major variability in West Africa, was evaluated for nine quantitative characters. The association between growth and yield parameters in this African plantain germplasm was examined to determine if the pattern of quantitative variation in inflorescence and vegetative traits agreed with taxonomic groupings based on inflorescence type and plant size. Phenotypic correlations between these traits were calculated. Giant cultivars were taller, their pseudostem thicker, and they flowered much later than medium-sized cultivars. Giant cultivars produced more foliage, resulting in heavier bunches with more hands and fruits. Groupings that resulted following principal component analysis (PCA) supported conventional taxonomic groupings of plantains. PCA was based mainly on time to flowering, pseudostem height, and number offruits. The last two traits, in combination with the number of hermaphrodite flowers and the persistence of the male bud, suffzced to group plantain cultivars.

Diversidad fenotipica y variaci6n en pl~tanos de Africa del Oeste y Central (Musa spp. grupo AAB Musaccae). Los pldtanos (Musa spp. grupo AAB) son un importante alimento y un componente integral de los sfstemas agrfcolas en los bosques tropicales h~medos del Africa del Oeste y Central. Una muestra de 24 cultivates de pl~tano, que representan la mayor variabilidad en Africa del Oeste, fue evaluada por nueve caracterfsticas cuantitativas. La asociaci6n entre las caracteristicas de crecimiento y rendimiento, en este germoplasma africano de pldtano, fue examinada para determinar si la variabilidadcuantitativa en la inflorescenciay los caractedsticos vegetativas estdn de acuerdo con grupos taxon6micos definidos por el tipo de la inflorescencia y el tamafzo de la planta. Correlacionesfenotipicas entre estas caracterfsticas cuantitativas fueron calculadas. Los cultivates gigantes son m~s altos, con un pseudotallo m ~ grueso y con floraci6n rods tardia que los cultivares de tama~o intermedio. Los cultivates gigantes producen m~sfollaje, 1o que resulta en rdcimos mds pesados con m~s manos y frutos. Los grupos definidos por el andlisis de componentes principales (PCA) estdn de acuerdo con grupos taxon6micos convencionales en pldtano. El PCA fue definido principalmente por dfas a la floraci6n, tama~o del pseudotallo, y n~mero de losfrutos por rdcimo. Estas dos (dtimas caractedsticas, en combinaci6n con el n~mero de flores hermafroditas y la persistencia de la bellota masculina, son suficientes par la clasifraci6n de cultivares de pl,~tano. Key Words: Plantains; Musa; Musaceae; Africa.

Musa, the Latin designation proposed by Linnaeus for the genus, is derived from the Arabic

' Received 21 June 1994; accepted 27 April 1995.

language (Simmonds 1966). The genus Musa

contains approximately 40 species with a center of origin in South East Asia (Simmonds 1962). Most cultivated Musa cultivars are triploid (2n = 3x = 33 chromosomes) perennial herbs de-

Economic Botany 49(3) pp. 320-327. 1995

9 1995, by The New York Botanical Garden, Bronx, NY 10458 U.S.A.

1995]

SWENNEN ET AL.: AFRICAN PLANTAINS

rived from interspecific crosses between the diploid species M. acuminata Colla. (A genome) and M. balbisiana Colla. (B genome) (Simmonds and Shepherd 1955). Plantains are almost completely sterile and develop fruit by parthenocarpy, i.e., without seeds because of the lack of fertilization. The role of Musa in agricultural development was highlighted by Jurion and Henry (1969), who suggested that edible bananas (Musa spp., AAA group), plantains (Musa spp., AAB group), and yams (Dioscorea spp.), provided the basis for agriculture in tropical Africa due to their vegetative propagation. Plantains and bananas are currently grown in 120 countries, which produce 75 l06 tonnes of fruit annually (FAO 1992). Ninety percent of the total production provides a significant source of domestic food for 400 million people in tropical developing countries. The remaining portion is exported to developed countries. Banana and plantain fruits are rich sources of carbohydrates, vitamins C, B, and A, and minerals such as K and Ca. While the consumption of banana in developed countries averages less than 30 g per person per day (i.e., one banana a week), the consumption in African countries may be more than 500 g per person per day (INIBAP 1989). Africa accounts for 35% of the world's plantain and banana production (INIBAP 1993). Plantains, which are produced in compound gardens or in fields in association with other food crops (Okigbo and Greenland 1976), have a high return on labor input (Nweke, Njoku, and Wilson 1988). Seventy million Africans obtain more than 10% of their calorie intake from banana or plantain fruits (Wilson 1987). In terms of annual gross value, plantain and banana rank first among other food crops such as maize, rice, and cassava in African agricultural production. The crop is primarily produced by smallholders, and is destined for both rural and urban markets. A clear ecoregional distinction is observed in the distribution of the cultivated types in subSaharan Africa. The AAB plantains are predominant in the humid lowlands of West and Central Africa, while AAA cooking and beer bananas prevail in the East African highlands (Swennen and Vuylsteke 1991). Although the crop originally came to Africa from Asia (Simmonds 1966), the wide array of cultivar variability found in Africa suggests a long and intense history of cultivation (De Langhe 1961, 1964). As such, the

321

humid forest zone of West and Central Africa is a secondary center of plantain diversification, while the East African highlands are a secondary center of diversity for bananas of the Musa A A A group. In most countries where plantains are a major staple, cassava or other crops account for a larger share of calorie intake. However, per capita consumption rates may reach as high as 150 kg per year in some traditional plantain-growing areas such as, for example, southwest Cameroon (Four6 and Lescot 1988). Although fruit is produced throughout the year, the major harvest comes in the dry season spanning the months of December through March (Nweke, Njoku, and Wilson 1988), when most other starchy staples are in short supply or difficult to harvest. Hence, plantain plays an important role in bridging the hunger gap (Wilson 1987). Fruits are slender and are usually cooked, roasted, or boiled before consumption because they are unpalatable when raw (Hahn, Vuylsteke, and Swennen 1990). Wide morphological variability is observed, particularly in inflorescence type, plant size, inflorescence and fruit orientation, fruit apex shape, and pseudostem and fruit colour (Swennen 1990; Tezenas du Montcel, De Langhe, and Swennen 1983). Plantain cultivars are grouped by their characteristic inflorescence type and the number and size of their fruits (Swennen and Vuylsteke 1987). Based on inflorescence morphology, four groups have been defined: French, French Horn, False Horn, and Horn plantains (Table l). Pseudostem height has also been used as a criterion for grouping plantains as giant, medium, or small (De Langhe 1964). Plant size reflects the number of foliage leaves produced prior to flowering, since leaf production stops at flowering. Giant plantains are those with more than 38 leaves, while small plantains have fewer than 32 leaves. Multivariate analysis techniques, such as principal component analysis (PCA) are more appropriate than univariate analyses when more than a single variable is measured on each individual (Iezzoni and Pritts 1991). While univariate analyses consider characteristics individually, PCA may reveal relationships among characters, such as vegetative and inflorescence characteristics, and/or determine how the West African plantain cultivars vary in terms of all variables considered together. This paper reports the pattern of variation in, and the association between, vegetative and inflorescence traits in

322

ECONOMIC BOTANY

[VOL. 49

TABLE l. PLANTAIN (MUSA SPP., A A B ) GROUPS ACCORDING TO THEIR INFLORESCENCE AND FRUIT CHARACTERISTICS.

Group

French

Inflorescence

Complete

Hermap,truditr flowers

Presence of male bud

Many

Yes

Number of hands*

6-10

Fruits Number

Many (70-130)**

French Horn

Incomplete

Many

No

7-8

False Horn

Incomplete

Few

No

5-11

Hom

Incomplete

None

No

1--4

Medium (41-80) Medium (23-70) Few (18--45)

Size

Small (0.15--0.25)*** Large (0.21--0.28) Large (0.26--0.46) Very large (0.29-0.45)

* Nodal clusters bearing fruits ** Number of fruits in parentheses (Swennen and Vuylsteke 1987). *** Fruit weight (kg) in parentheses (Swennen and Vuylsteke 1987).

the West and Central African plantain germplasm, using principal component analysis. MATERIALS AND METHODS Twenty-four plantain cultivars were selected for this research (Table 2). These cultivars constitute a representative sample of the total plantain variability in West Africa (Swennen and Vuylsteke 1987), where 116 cultivars have been described (Swennen 1990). These 24 plantains were originally collected in C6te d'Ivoire, Ghana, Nigeria, and Cameroon but are also cultivated in other African countries under different names. Experiments were conducted in 1983-1984 at the High Rainfall Station of the International Institute of Tropical Agriculture, Onne (4"43'N, 7*0 I'E, southeastern Nigeria), which is in the secondary center of plantain diversity. Annual rainfall averaged 2400 mm, with a monomodal distribution from February through December. Average temperatures ranged from 25~ in July to 27"C in February, March, and April. Relative humidity was high, ranging from 78% (February) to 89% (July and September). Average sunshine per day was 4 hours, ranging from 2 h (September) to 6 h (February). The soil was a Typic Paleudult, derived from the coastal plain sediments of the Niger Delta region. It was deep and well drained, but highly acidic (pH 4.2) and deficient in nutrients. The plantain comas were planted in holes 3040 cm deep. Each plot consisted of five plants in a row. Row spacing between cultivars was 3 m, and plants of the same cultivar were separated within rows by 2 m. The field was mulched with cut Pennisetum purpureum at a rate of 80 t ha-

year-~ (fresh weight). Total fertilizer rate per ha was 300 kg N and 456 kg K, applied in six equal applications during the rainy season. The following inflorescence and vegetative characters were measured: bunch weight (kg), number of nodal clusters bearing fruits (or hands) per bunch, number of fruits per bunch, fruit weight (kg), pseudostem height (cm) at flowering, pseudostem circumference (cm) at 50 cm above soil level at flowering, time to flowering (number of days between date of planting and date of bunch emergence), number of foliage leaves produced until flowering, and length/width ratio of the seventh youngest leaf at flowering. Data were recorded as described by Swennen and De Langhe (1985). Phenotypic correlations were calculated for the four inflorescence and five vegetative traits. Principal component analysis (PCA) was conducted using the traits evaluated to reveal patterns within the data matrix. PCA was performed on the 24 • 9 plantain • trait mean matrix using MSTAT-C software (Anonymous 1989). First and second principal components (PRIN) were plotted to enhance the dispersion of the 24 plantain cultivars, based on their inflorescence and vegetative characteristics. RESULTS AND DISCUSSION The matrix of correlations between inflorescence and vegetative traits of plantain (Table 3) indicated that bunch weight was significantly affected by the number of hands and fruits per bunch, but not by fruit weight. The lack of a significant correlation between fruit weight and bunch weight appeared to contradict the gener-

1995]

S W E N N E N ET AL.: A F R I C A N P L A N T A I N S

323

z 0

o

E 0

E

E 0E0~~E0E 0

0

z

<

z

5 o

9

o

I =1

m

~o~

I-

"

zm

9

m

<

0

0

i

0 f't n.

o I=

0 0

=o

.=

r,..)

0

e~ e~ z < 0

0

~-~

o

.~ , . . . ~ ~

~

.o ~ ~ ~,_,

.~

~ ~~ ~ , . . ~ ~.-

.

0 ~.~:~

~ z I",

i

,.~

o .~

= -= o a "

o

'=

I=

o

~

=..-,-~ ~ :~

,=

324

ECONOMIC BOTANY

[VOL. 49

TABLE 3. PHENOTYPIC CORRELATIONS BETWEEN BUNCH AND VEGETATIVE TRAITS IN W E S T AFRICAN PLANTAIN(MusA SPP., A A B GROUP) GERMPLASM. Tratt

2

1. Bunch weight

3

0.556 0.887 (0.004) (0.000) 0.490 (0.013)

2. No. of hands 3. No. of fruits 4. Average fruit weight

4

-0.316 (0.123) -0.398 (0.048) -0.652 (0.000)

AND CENTRAL

5

6

7

8

9

0.473 (0.017) 0.380 (0.060) 0.363 (0.074) 0.141 (0.501)

0.782 (0.000) 0.336 (0.100) 0.578 (0.002) 0.119 (0.571) 0.802 (0.000)

0.801 (0.000) 0.450 (0.027) 0.603 (0.002) 0.087 (0.686) 0.802 (0.000) 0.918 (0.000)

0.752 (0.000) 0.403 (0.045) 0.520 (0.008) 0.166 (0.429) 0.737 (0.000) 0.912 (0.000) 0.943 (0.000)

0.086 (0.690) 0.096 (0.656) 0.103 (0.673) 0.070 (0.746) 0.745 (0.000) 0.357 (0.086) 0.368 (0.084) 0.452 (0.026)

5. Pseudostem height 6. Pscudostem circumference 7. Time to flowering 8. Number of leaves 9. Leaf length/width ratio Slgmficance level, a, for false rejection of null hypothesis: p = O, is indicated in brackets.

ally accepted hypothesis that when bunch weight decreases, the average fruit weight increases, from the French through the French H o r n and False H o r n inflorescence type to the most degenerated H o r n type (Tezenas du Montcel and Devos 1978; Tezenas du Montcel, De Langhe, and Swennen 1983). Fruit weight depended on both number o f fruits and hands per bunch but it was not correlated with the other traits. The n u m b e r o f leaves produced prior to flowering was correlated with all the traits except fruit weight and leaf length/width ratio. Tall plantain cuitivars had a larger girth at flowering than short TABLE 4.

E I G E N VECTOR VALUES FOR PRINCIPAL

COMPONENTS USING BUNCH AND VEGETATIVE TRAITS IN W E S T AND C E N T R A L A F R I C A N PLANTAINS

(MusA s P a . ,

AAB

GROUP).

Trait

PRIN 1

PRIN2

Bunch weight No. of hands No. of fruits Average fruit weight Pseudostem height Pseudostem circumference Time to flowering Number of leaves Leaf length/width ratio % of total variation

0.087 0.015 0.352 0.000 0.537 0.145 0.746 0.057 0.002 79.409

0.123 0.016 0.850 -0.002 -0.510 0.003 -0.049 0.005 -0.002 14.239

cultivars (r = 0.802; P < 0.001). The highly significant correlation between pseudostem height and time to flowering (r = 0.802; P < 0.001) indicates that tall plantains flowered later than intermediate height plantains and intermediate height plantains flowered later than short plantains. Similarly, tall plantains had more and narrower leaves than short plantains (r = 0.737, P < 0.001; and r = 0.745, P < 0.001, respectively). All vegetative characters, except leaf length/ width ratio, were correlated with bunch weight. High bunch weights were positively correlated with pseudostem height and circumference (r = 0.473, P < 0.05 and r = 0.782, P < 0.001, respectively), and n u m b e r o f leaves (r = 0.752; P < 0.001). Also, changes in bunch weights were positively correlated with increases in time to flowering (r = 0.801; P < 0.001). This indicated that giant cultivars were taller, with a thicker pseudostem and flowered later than intermediate height cuitivars, and produced more foliage, resuiting in heavier bunches. PCA reduced the nine original variables (inflorescence and vegetative characteristics) into two major components that together accounted for 94% o f the original variation (Table 4). The first and most important component P R I N 1 accounted for 79.4% o f the total variation. The interpretation o f PRIN1 was facilitated by examination o f the eigenvector values, which are

1995]

SWENNEN ET AL.: AFRICAN PLANTAINS

325

_

7.

O

French

[]

French Horn

O -100

False H o r n

o

OOoO

==

13

Horn

o

[] []

o

o -200

I

300

I

!

400

,

600

500

PRIN

700

1

Fig. 1. Plot of the first (PRIN 1) and second (PRIN2) principal components from analysis of vegetative and bunch traits of West and Central African plantains. Symbols based on their bunch type. the relative weights accounted for by each individual variable. Inflorescence and vegetative characteristics contributing to P R I N 1 were time to flowering and pseudostem height (Table 4). Late flowering, giant size, and many fruits also contributed to PRIN1 variation. Thus, giant French plantains (Fig. 1, 2) had high PRIN1 scores, while small plantains had a low P R I N 1 score (Fig. 2). The second component (PRIN2)

accounted for a smaller portion o f the total variation (14.2%). Variation in P R I N 2 was primarily accounted for by number o f fruits (positive correlation) and a negative correlation with pseudostem height. Hence, medium French plantain had high PRIN2 scores, while Horn or False Horn plantains had low P R I N 2 scores (as indicated in Fig. 1,2). The plantain cultivars were grouped mainly in

_

A 79 .9

& -100 9

A

#~

A

A

AA -200

, 300

400

A

/k

Giant

A

Medium

9

Small

Lx A

A

, 500

PRIN

,

600

, 700

1

Fig. 2. Plot of the first (PRIN1) and second (PRIN2) principal components from analysis of vegetative and bunch traits of West and Central African plantains. Symbols according to their plant stature.

326

ECONOMIC BOTANY

three classes according to the pattern o f variation, as determined by PCA, which reflected their inflorescence type: French, French Horn, and False/True Horn. The French Horn cultivars were distributed between the French and Horn (False and True) plantains (Fig. 1). This pattern o f variation could reflect the relationship between the plantain cultivars and their c o m m o n evolutionary pathway from French to H o r n types. The plantains were clearly subdivided by PCA according to their sizes (Fig. 2). Two distinct subgroups were evident in the French plantains: m e d i u m and giant (Fig. 1 left and right, respectively). Similarly the False H o r n and the French H o r n plantains were subdivided in giant, medium, and small plantains according to their sizes (Fig. 1, right to left). The False and True Horn plantains were not differentiated by the PCA, probably because all the True H o r n plantains are o f m e d i u m size. The PCA groupings and subgroupings (Fig. 1, 2) support the conventional morphological taxo n o m y o f plantain based on inflorescence type and plant stature. Pseudostem height and number o f fruits were the characteristics that simult a n e o u s l y c o n t r i b u t e d m o r e to P R I N 1 a n d PRIN2. Therefore, both traits in combination should be used as discriminant characters to assign plantain cultivars to each taxonomic group. Thus, four m a j o r plantain groups based on their inflorescence type, and three m i n o r subgroups within each major group based on their plant stature (except for m e d i u m True H o r n plantains) are defined. These two quantitative traits and traits such as the number o f hermaphrodite flowers and the persistence o f the male bud in the inflorescence axis (Table 1), suffices to group plantain cultivars. The principal component analysis also suggests that other traits (e.g., fruit size or number o f leaves) normally used by taxonomists are not required for plantain classification due to their significant correlation with either number o f fruits or pseudostem height (Table 3). Time to flowering, which was an important contribution to P R I N I , was positively correlated with pseudostem height (r = 0.802, P < 0.001), and number o f frnits (r = 0.603, P < 0.01). LITERATURE CITED Anonymous. 1989. MSTAT-C. A microcomputer program for the design, management and analysis

[VOL. 49

of agronomic research experiments. Michigan State University, East Lansing, MI. De Langhe, E. 1961. La taxonomie du bananier plantain en Afrique Equatoriale. Journal d'Agriculture Tropicale et de Botanique Appliquee (Brussels) 8:419-449. 9 1964. The origin of variation in the plantain banana. Mededelingen Landbouwhogeschool Gent 29:45-80. FAO (United Nations Food & Agriculture Organization). 1992. FAO Yearbook, Production 1991, Vol. 45. FAO, Rome, Italy. Fourt, E , and Th. Lescot. 1988. Variabilit6 g6n6tiquc des Mycosphaerella inf6od6s au genre Musa. Mise en 6vidence de la pr6sencc au Cameroun sur bananiers et plantains d'une cercosporiose (Mycosphaerella musicola) au comportement pathog6ne atypique. Fruits 43:407-4 15. Hahn, S. K., D. Vuylsteke, and R. Swennen. 1990. First reactions to ABB cooking bananas distributed in southeastern Nigeria. Pages 306--315 in R. A. Fullerton and R. H. Stover, eds., Sigatoka leaf spot diseases of bananas. Proceedings of an international workshop held at San Jose, Costa Rica, 28 March1 April, 1989. INIBAP, Montpellier, France. lezonni, A. F, and M. P. Pritts. 1991. Applications of principal component analysis to horticultural research. HortScience 26:334-338. INIBAP (International Network for the Improvement of Banana and Plantain). 1989. Annual Report. International Network for the Improvement of Banana and Plantain, Montpellier, France. 9 1993. Medium-Term Plan 1994-1998. International Network for the Improvement of Banana and Plantain, Montpellier, France. Jurion, F., and J. Henry. 1969. Can primitive farming be modernised? Hors Series, Belgian Cooperative and Development Office. Brussels, Belgium. Nweke, F., J. Njokn, and G. F. Wilson. 1988. Productivity and limitations of plantain (Musa spp. cv. AAB) production in compound gardens in Southeastern Nigeria. Fruits 43:161-166. Okigbo, B. N., and D. J. Greenland. 1976. Intercropping systems in tropical Africa. Pages 63-101 in R. Paperdick, P. Sanchez, and G. Tripplet, eds., Multiple cropping. Spec. Pub. No. 27, American Society of Agronomy, Madison, WI. Simmonds, N . W . 1962. The evolution of the bananas. Tropical Science Series. Longman, London. --. 1966. Bananas. 2nd ed. Tropical Agricultural Series. Longrnan, London. , and K. Shepherd. 1955. Taxonomy and origins of cultivated bananas. The Journal of the Linnean Society of London, Botany 55:302-312. Swennen, R. 1990. Limits of morphotaxonomy: names and synonyms of plantain in Africa and elsewhere. Pages 172-210 in R. L. Jarret, ed., Identification of genetic diversity in the genus Musa. Pro-

1995]

SWENNEN ET AL.: AFRICAN PLANTAINS

ceedings of an international workshop held at Los Bafios, Philippines, 5-10 September 1988. INIBAP, Montpellier, France. - - , and E. De Langhe. 1985. Growth parameters of yield of plantain (Musa cv. AAB). Annals of Botany 56:197-204. , and D. Vnylsteke. 1987. Morphological taxonomy of plantain (Musa cultivars AAB) in West Africa. Pages 165-171 in G. J. Persley and E. A. De Langhe, eds., Banana and plantain breeding strategies. Proceedings of a workshop held at Cairns, Australia, 13-17 Oct 1986. ACIAR Proceedings No. 21. , and . 1991. Bananas in Africa: diversity, uses and prospects for improvement. Pages 151-159 in N. Q. Ng, P. Perrino, F. Attere, and H.

327

Zedan, eds., Crop genetic resources of Afriea, Vol. II. Proceedings of an international conference held at Ibadan (Nigeria), 17-20 Oct 1988. IITA/IBPGR/ UNEP. Tezenas de Montcel, H., and P. Devos. 1978. Proposal for establishing a plantain determination card. Paradisiaea 3:14-17. , E. De Langhe, and R. Swennen. 1983. Essai de classification des bananiers plantains (AAB). Fruits 38:461-474. Wilson, G . F . 1987. Status of bananas and plantains in West Africa. Pages 29-35 in G. J. Persley and E. A. De Langhe, eds., Banana and plantain breeding strategies. Proceedings of a workshop held at Cairns, Australia, 13-17 Oct 1986. ACIAR Proceedings No. 21.

BOOK REVIEW Flora Mesuamericana. Volume 6. Alismataceae a Cyperaceae. G. Davidse, M. Sousa S., and A. O. Chater, general editors. 1994. Universidad Nacional Aut6noma de M6xico, Missouri Botanical Garden, and The Natural History Museum (London). xvi + 543 pp. (hardcover). (Department Eleven, Missouri Botanical Garden, P.O. Box 299, St. Louis, MO 63166.) $75.00. ISBN 968-36-3310-2. Botanists of every stripe will applaud the appearance of this first volume (ofa projected 7) of a flora estimated to contain 18 000 species of vascular plants. The reader is struck immediately by the resemblance to Flora Europaea, right down to the design of the dust cover. The editors could not have chosen a better model. The improvements on the model are welcome ones: no Roman numerals are used in the running heads; voucher specimens are cited in support of the range statements; more synonymy is given; nomenclatural types are cited; the descriptions are about twice as long; and Spanish c o m m o n names are given here and there (not feasible in polyglot Europe, of course). One imagines there were anguished discussions over what language to use for the flora. The decision to adopt Spanish, the first language of the entire area except Belize, where Spanish is a widely used second language,was a good one. My impression is that the translators made an effort to keep the Spanish simple; I still needed a dictionary here and there, but in general I found I could get through the [wholly artificial] keys and ample descriptions fairly easily. The language choice is intended to make the work accessible to local hot-

anists in the area covered; the price of course works contrary to this noble motive. The Flora project was formally initiated in 1980, which might imply that some of the treatments are very much dated. This seems not to be so: for example, a paper on certain species of Paspalum, some of which occur in Middle America, appeared 1 March 1995 (Annals Missouri Botanical Garden 82(1):82-116). For species c o m m o n to the two treatments, I judge the Flora descriptions are entirely satisfactory, and only extralimital range statements require amplification (e.g., Paspalum reclinatum Chase, page 340, occurs not just in Costa Rica and Colombia, but also in Ecuador). The sequence of flowering plant families follows Melchior, Syllabus der Pflanzenfamilien, 1964; the families in volume 6 are numbers 234 through 263, while numbers 264 through 274 will appear in volume 7. However, all families of monocots are covered in the introductory keys, pp. 1-2; the families yet to be treated are not indicated, which might have been helpful. The families not in the present volume include inter alia Arecaceae (Palmae), Araceae, Musaceae, Marantaceae, and Orehidaceae. The next volume to appear (volume 1) will cover ferns and fern allies. The editors hope to have future volumes appear at approximately two-year intervals. They will be eagerly awaited. NVJL A. BIOLOGY D E P A t t ~ UNIVERSITY OF WISCONSIN-OSHKOSH

OStmOSH, WI 54901