Plant Systematics and Evolution ? Springer-Verlas 1999

Pl. Syst. Evol. 216: 23–47 (1999)

Printed in Austria

Pollination mechanisms and pollen-ovule ratios in some Genisteae (Fabaceae) from Southwestern Europe JOSEFA LÓPEZ, TOMÁS RODRÍGUEZ -RIAÑO,

ANA

ORTEGA-OLIVENCIA,

JUAN ANTONIO

DEVESA, and TRINIDAD RUIZ Received May 8, 1997; in revised version April 9, 1998

Key words: Leguminosae, Genisteae. – Spain, nectar, pollen, pollination mechanisms, pollen/ovule ratios, anther biomass. Abstract: We studied the biology and floral rewards of 34 taxa of Genisteae from the SW of Europe. Most of the floral attributes show a statistically significant direct relationship. Anther biomass of the lower whorl (1W) is significantly lower than that of the upper whorl (uW), and the ratio of the two (Rv) differs among the taxa. All taxa are polliniferous, and Retama sphaerocarpa also secretes nectar. They can be considered xenogamous or facultative xenogams on the basis of their high pollen/ovule (P/O) ratios. Three principal mechanisms of pollen releasing are identified in this tribe: valvular, pump and explosive; the latter comprises specialized and nonspecialized subtypes. Pollination is sternotribic except in the specialized explosive mechanism, in which it is noto-sternotribic. Thus some pollen serves as food (pollen from the uW, which adheres to the ventral surface of the insect) and part of the pollen fulfils a polliniferous function (pollen from the 1W, which adheres to the dorsal surface). Species that use a pump mechanism have very low Rv values (Rv=0.08–0.26); species with valvular or nonspecialized explosive mechanisms have Rv values between 0.24 and 0.58; those with a specialized mechanism of pollen presentation have high Rv values (0.6–0.76). In contrast to expectations, the highest P/O ratios appear in the specialized explosive system, which allows a single visit from the insect. The tribe Genisteae (ADANS.) BENTH. comprises 14 genera according to POLHILL (1976) and up to 20 according to BISBY (1981). Representatives are present in the Mediterranean region, and throughout Europe and the Canary Islands, and several taxa are located in South Africa, Madagascar and India. Of these genera Lupinus L. is the only one that is widely distributed in the New World. This is the most diverse genus of the group, with nearly 200 species, whereas the rest of the tribe comprises about 273 species (BtsBY 1981). Most members of the tribe are shrubby or subshrubby, and many species of Lupinus are herbaceous. Lupinus is considered a natural group, distinguishable from the rest of the tribe by its morphology, pollination mechanism, root nodules, biogeography, and (to a more limited extent) chemical composition. These features

24

J. LópEZ & al.:

led BISBY (1981) to recognize two subtribes within the tribe Genisteae: Lupininae and Genistinae. The Genisteae have typical butterfly-like zygomorphic flowers that are mostly adapted to bee pollination (MEEUSE 1961, FAEGRI & VAN dER PIJL 1979). The different parts of the corolla play an important role with a view to the flower functions. The standard helps to attract the pollinator, the keel protects the staminal column and together with the wings provides a platform for the insects to land on. The wings also operate as levers that raise or lower the carina, and give rise to an attractive structure (ARROYO 1981, STIRTON 1981). The petals are commonly yellow, and less frequently whitish, bluish or purple (Pout-nu 1976). Some petals reflect UV radiation either because of pigment variation or the presence of reflecting hairs (BisBY 1981). Therefore the appearance of the flowers to insects is variable, the greatest differences between genera and species depending on the wing characteristics (KAY 1987). In Lupinus, variegated corollas are frequent, and the contrast may be even more evident after pollination (e.g. L. micranthus Guss.; L. angustifolius L., BISBY 1981; and L. argenteus PURSH, GoRI 1989), a feature that has been noted previously for the genus Chamaecytisus LINK. The most significant functional and taxonomic characteristic of Genisteae is the presence of a typical monadelphous androecium (POLHILL 1976). Unless extrastaminal nectaries exist or autogamy or deception is present, the main attraction of Genisteae flowers to insects is their copious pollen production (Poulin, 1976). In Fabaceae four principal mechanisms of pollen presentation have been recognized: pump, valvular, brush and explosive. Whereas in the first three the flowers can receive several visits, in the last a single visit is effective (DELPINO 1868, LEPPIK 1966). According to LEPPIK (1966) the papilionaceous flower has evolved to greater pollen economy, reaching its biggest success with tripping mechanisms. According to ARROYO (1981), knowledge of the pollen-ovule ratio (P/O), little studied in this family, is useful for detecting possible evolutionary trends in pollen economy. If explosive pollen delivery is the most specialized mechanism (ARROYO 1981), those species with this mechanism would be expected to have low P/O ratios. In the genus Medicago L., whose flowers use an explosive mechanism, P/O ratios are lower than in related genera of the tribe Trifolieae BENTHAM (SMALL 1988). Very little is known of nectar and pollen production in this group (but see HERRERA 1985, 1987; WEBB & SHAND 1985) or of P/O ratios of species of this family. Typifying and quantifying floral rewards in each taxon is of great importance from an applied (e.g. apicultural) point of view, and would shed light on pollinator affinities, whereas knowledge of the P/O ratios provides indirect information about the breeding system. Except for the genus Lupinus (HoRovITz & HARDING 1972, SCHAAL & LEVERICI 1980, SCHAAL 1989) and Chamaecytisus (WEBB & SHAND 1985), no published information exists for the Genistinae. In view of the foregoing, the present study was designed with the following objectives: (1) to elucidate the relationships among the different floral attributes of Genisteae; (2) to typify and quantify the floral rewards (pollen and nectar) offered

Floral biology in Genisteae

25

to pollinators, and (3) to determine the P/O ratios of the species, and whether these ratios differ in relation to pollen presentation mechanisms. The work focuses on the Genisteae in the flora of Extremadura (SW Spain), where 36 taxa have been cited (DEVESA 1995). Many of the taxa are important components of Mediterranean forests and scrubs, and are also of interest for apicultural or farming purposes. Better knowledge of their floral and reproductive biology will aid in the management and conservation of these natural ecosystems. Materials and methods Morphology and floral biometry. We studied a total of 88 wild populations from the SW of Spain, belonging to 34 taxa (see Appendix 1). An average of three populations for each taxon and ten flowers per population were studied. Vouchers are conserved in the University of Extremadura Herbarium (UNEX). In each population the colour of the flowers was determined by visual observation in fresh material. Floral size was measured as the maximum length of the flower on its vexilar side, from the base of the calyx to the upper part of the standard in dried material. In dissected flowers we measured ovary and style length and determined the morphology of styles and stigmas. Ovule production per flower was calculated from floral buds preserved in 70% ethanol. Other floral buds were used to study biometric and morphological characteristics of the androecium. Anther size was obtained from two anthers per verticil (B2, D6 from the upper and C4, E8 from the lower whorl; see Fig. 1), and from the F10 or Al anthers when they differed greatly in size from the rest. These anthers were moistened, mounted on a slide and examined with a light microscope. For each anther maximum length (L) and width (W) were obtained, and these values were used to calculate volume. Because the anthers of both whorls were dimorphic, two geometric figures were considered for each flower: (a) a cylinder for the upper whorl (uW), with the formula V= RA, W 2 )/2]/4 (thickness nearly 1/2 the width) and (b) either a cylinder

Fig. 1. Floral diagram of a Genisteae. In the monadelphous androecium, the large ovals represent upper whorl anthers, and the small ones lower whorl anthers

26

J.

LÓPEZ

& al.:

or a cylinder and a cone for the lower whorl (1W), with the formula V total = Vcylinder + Vcone = [(piL W2)/2]/4 + [(pi L W 2 )/2]/12. For each taxon the following attributes were also calculated: (1) biomass of the anthers of each whorl (as the average size of the anthers multiplied by the number of anthers); (2) total volume of the whole anthers of the flower (total biomass), and (3) Rv, obtained by dividing 1W anther biomass by uW anther biomass. Pollen size was also studied in acetolysed grains from 89 populations (average of three populations per taxon) with the method of ERDTMAN (1960) modified by HIDEUx (1972). From each slide 15 pollen grains were studied and their volumes were calculated based on the formula for a revolution ellipsoid, 4/3 ra2b, where a is one half of the equatorial axis and b one half of the polar axis. Floral rewards. Nectar. An average of three populations per taxon and 15 flowers per population were studied. Flowers were obtained from flowering branches previously bagged in the field, and were analyzed after 24 h. None of the species produced nectar except Retama sphaerocarpa, although it was reabsorbed or evaporated. Nectar production was estimated in this species in the field in flowering branches from different individuals that had been bagged in nylon nets for 24 h. Production was quantified at 08.00 h (solar time) by measuring the volume with 0.5 µl micropipets, and concentration was measured with two pocket refractometers adapted to small volumes (Bellingham and Stanley Ltd., 0–50% and 40-85%). P o 11 e n . The total production of pollen grains per flower was calculated for the same floral buds as were used to study the ovules. Normally a total of five floral buds per population were studied, or 15 when only one population was available. Direct counting was used when the approximate number of pollen grains per flower was less than 2000, and the dilution method (CRUDEN 1977) when more than 2000 grains per flower were present. For direct counting anthers Al, B2, C4, D6, E8 and F10 (Fig. 1) were each placed in a drop of lactophenol blue on a slide, and squashed gently to spread the pollen. Grains were counted with the help of an eyepiece grid and a manual counter. Production of the uW was calculated with the formula uWP=2(B2+D6)+F10, and that of the 1W with the formula 1WP=A1+2(C4+E8). Total pollen production per flower was considered the sum of the values from the two whorls. With the dilution method the anthers of the two verticils were considered separately in most cases (e.g. Cytisus L. and Adenocarpus DC.), but in some cases (e.g. Lupinus) the 1W anthers were quantified by direct counting while the uW anthers were studied with the dilution method. In all counts those pollen grains that were completely stained with lactophenol blue (fertile) were distinguished from collapsed or weakly stained grains (sterile). Pollen biomass per flower was calculated by multiplying the average number of pollen grains per flower by the average size of the grains. Pollen biomass per verticil was also calculated. Tripping mechanisms and pollen-ovule ratios. Because previous studies deal with only a few species (MULLER 1883, PROCTOR & al. 1996, among others) direct observations were made both in the field and on dried material to determine the mechanism of pollen releasing. Pollen (P) and ovule (0) production per flower were used to calculate the P/O ratio, and these ratios were used to classify the taxa according to the categories proposed by CRUDEN (1977). Statistical analysis. The statistical packages Systat for Windows (Vers. 5; Systat 1992) and Statgraphics for Windows (Vers. 1.0; Statgraphics 1994) were used. A correlation matrix was made using mean values of the floral attributes for each taxon, and these values were subjected to Spearman's rank correlation test. Significance of the values was

Floral biology in Genisteae

27

determined with the sequential Bonferroni test at a <0.05 (Rica 1989). To find possible differences between groups of taxa we used the Mann-Whitney and Kruskal-Wallis nonparametric tests (SOKAT & ROHLF 1979).

Results Morphology and floral biometry. The largest floral lengths in the taxa (given as means ± standard deviation, S.D.) were found in Spartium junceum L. (26.08 ± 2.36 mm) and Cytisus grandiflorus (BROT.) DC. (24.10 ± 0.99 mm), and the smallest ones in Retama sphaerocarpa (4.38 ± 0.39 mm) and Genista anglica L. (7.42 ± 0.27 mm) (Table 1). Yellow was the most frequent colour of the flowers (82.35% for the taxa), whereas whitish, bluish-white and violet-purple flowers appeared in a minority (approx. 5.88% for each color). The largest ovary sizes were found in the flowers of Adenocarpus hispanicus subsp. gredensis RIVAS MARTÍNEZ & BELMONTE and Spartium junceum, and the smallest in Retama RAFIN. The longest styles appeared in Cytisus grandiflorus and C. striatus (HILL) ROTHM., and the shortest in Retama sphaerocarpa and Argyrolobium zanonii (TURRA) P. W. BALL (Table 1). Styles were generally linear in shape, appearing straight or slightly curved, either to the apex (e.g. Genista L.), or to the base (e.g. Lupinus and Adenocarpus). This curvature was especially evident in the styles of Cytisus arboreus subsp. baeticus ( WEBB) MAIRE and C. scoparius (L.) LINK. In some cases the style became circinate after pollination, as in Cytisus striatus and C. grandiflorus. They were usually glabrous, although occasionally a small pubescence appeared either on the dorsal part or at the base. In these Cytisus species the apical part of the style was bent such that the stigma nodded, and at the point of flexure the style was thickened, and a slight ventral hollow was seen. This cavity was well developed in Cytisus striatus, C. scoparius and C. grandiflorus, but was absent in C. arboreus. Style length correlated directly with floral size and ovary size (Table 3); i.e. large flowers had long styles, and as the ovaries became larger, so did the styles. The stigma was variable in the tribe. It was generally papillate, with marginal papillae usually longer than the central ones. Morphologically, the stigma was classified as any of the following types: (1) terminal and globose, either crowned (sterile hairs all around) or acrowned; (2) tufted; or (3) discoid. Within the latter two subtypes were recognizable: (a) erect and terminal (apical location) and (b) lateral and introrse. Some stigmas were (4) rectangular-lateral and introrse. Figure 2 and Table 1 summarize the types of stigmas in the different taxa. In terms of the number of ovules per flower (Table 1), the most productive taxa were Genista falcata BROT. (22.60 ± 2.67) and G. anglica (17.90 + 1.10), and the least productive were Lupinus micranthus (2.80+0.45) and L. luteus L. (3.80 ± 0.56). This variable correlated directly with floral size and the ovary size, although in the former case the correlation was not significant according to the sequential Bonferroni test (Table 3). All taxa had a monadelphous androecium, and the free part of the filaments was linear. The anthers, dimorphic in all cases, were basifixed in the uW and

Y

Y

Y

Y

Y

Y

Y

G. /tot-ida

G. hirsula

G. pol-yonthos subsp. hvslrix G.laurntfaTii

G. lriacanlhos

G. lridenlala

G. mnhellala

R. monasperma

Resana

W

Y

G. faleala

average

Y

G. rinerasvrns

Genista,

Y

Flower colour

G. anglirn

Genista

Maxon

10.0010.41 (10)4.08

5.2010.16 ( I (ï)á.04

(30) 8.28 6.7210.45 (30) 6.68 6.62]1158 (30) 8.80 11.10:131.52 (30) 4.64 6.151058 (10) 9.43 6.3010.26 (10) 4.10 6.4310.41 (30) 6.37 8.8710.86 (30) 9.70 8.25+0.35 (10) 4.29 7.49+1.81 (222) 24.16

(30) 8.05 5.3010.47 (30) 8.79 5.02 1.0.3 I (30) 6.13 2.00.1_-0.00 (30) 0.00 4.2010.42 (10) 10.04 3.051:0.16 (10) 5.18 2.0410.35 (30) 17.34 4.0310.57 (30) 14.17 3.1510.24 (10) 7.67 3.6711.27 (222) 34.62

1.601(1.12 (10)7.44

4.851.0.38 (12) 7.76 6.8210.57

111111

Style size

2.79.1.0-28 (12) 10.05 4.221034

111111

111111

7.42.1 0.27 (12) 3.62 11.42 I 0.54 (30) 4.75 11.5810.77 (30) 6.62 12.25. I 0S5 (30) 4.52 13.15. L0.72 (30) 5.48 9.90J 0.74 (10) 7.45 9.8510.24 (10) 2.45 8.2310.48 (30) 5.89 12.67111)6 (30) 8.38 12.3510.63 (I))) 5.07 11.2111.91 (222) 16.99

Ovary size

Flower size

Gynoccium

SC

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

Style type

'r

D'I'

T

DI

DI

DI

DI

T

DI

'I'

DI

Stigma type

4.40+0.89 ©20.33

(15) 5.12 22.6012.67 (15) 11.80 6.801056 (15) 8.24 6.47 (1.74 (15) 11.49 7_4011.14 (5) 15.41 3.8010.84 (5) 22.02 4.2710.70 (15) 16.49 4.2710.70 (15) 16.49 4.60+0.51 (15) 11.02 8.5816.38 (125) 74.29

17.90- 1_ 1.10 (10) 6.15 6.82.10.35

Ovule number

1W-D

uW-B-L

IW- D

uW-B-L

IW- D

uW-B-L

1W-D

uW-B-1.,

1W- D

uW-B-L

IW-D

uW-B-L

1W-D

uW~l.

1W-D

uW-B-I,

1W-D

uW-B -L.

I W-D

uW-B-1,

I W-D-O

uW-B- 1,

Anther type

61.84111.02 (IO) 17.82

I77.42.1 59.25 (20) 33.40 126.85125.68 (27) 20.24 203.02153.82 (30) 26.51 156.77.1_28.28 (30) 18.1)4 238.17.183.53 (30) 35.07 151.33137.37 ( I 0) 24.69 136.08111.28 (11)) 8.29 76.35.1..12.46 (30) 16.32 212.56146.54 (23) 21.90 207.44127.46 (10) 13.24 168.36168.51 (2201) 40.69

j

Anther

dIt1111

size

uW

(1 0) 10.55

19.8712.10

(27) 17.50 59.25125.44 (30) 42.93 63.69117.06 (30) 26.78 87.63 1.34.96 (30) 39.89 42.55.1.16.46 (10) 38.68 56.531.5.97 (10) 10.56 31.9614.32 (30) 13.52 54.59.18.32 (23) 15.24 91.98111.49 (10) 12.49 55.7I +2672 (220) 47.95

44.01 119.6(1 (20) 44.54 34.641 6.06

1W Anther size Oman '

And race) um =

0.331007 (10) 22.34

(20) 17.51 0.29 . LO.08 (27) 28.59 0.2910. (30) 36.85 0.40.1.0.07 (30) 17.23 0.37.1 0.08 (30) 22.09 0.28 I (1.(16 (10) 21).25 0.42.1.0.04 (10) 8.55 0.43.L0.07 (30) 15.37 0.27d,0.07 (23) 27.49 0.4510.07 (10) 15.56 0.3410. I0 (220) 29.84

(1.2410.(14

I W/nW

Kv

'

390.78

408.51153.18 (10) 13.02

(20)'' 35.30 807 451 132.10 (27) 16.36 13 1 1.301350.67 (30) 26.74 1102.27.1.21525 (30) 19.53 1628.97.1.56098 (30) 34.44 969391 264.47 (10) 27.28 963.031.78.04 (10) 8.10 54156]73.91 (30) 13.65 1335.79+24.0(1 (23) 18.19 1497.091161.46 (IO) 10.79 1 12(1.391449.72 (220) 40.14

11(17.14

dinin

Anther biomass/II

8.5911.34 (15) 15.58

9.25 1.1.05 (30) 11.37 1038 1 35 (45) 12.96 10. 16.1..1.20 (45) 11.80 7.75 (45) 18.25 2229 13.55 (45) 15.93 9.231.0.80 (15) 8.72 11.171_1.70 (15) 15.21 8.89 (45) 9.94 213313.90 (45) 1830 7.0610.66 (15) 9.38 12.5416.00 (345) 47.88

Pullen size pm '

Table 1. Qualitative and quantitative characteristics of the flower in Genisteae. For each (axon quantitative characters show the mean value ± standard deviation (first line), and the sample size in parenthesis and the coefficient of variation (second line). Flower colour: Y yellow, W white, BW whitish-blue, VP violet-purple. Style types: SC straight and slightly curved, SA straight and curved at the apex, SB straight and curved at the base, StC strongly curved, C circinate. Stigma types: GA globose-acrowned, GC globose-crowned, DT discoid-terminal, Dl discoid-introrse, T tufted, R rectangular-lateral. Morphology of the anthers: uW upper whorl, IW lower whorl, B basifixed anthers, D dorsifixed anthers; L linearoblong, LA linear-oblong acuminate, O orbicular. Rv is the 1W/uW anther biomass ratio

sphaemcarpa

Y

U. minor

Y

granrAJlorus

muhiflarus

scopaa-ius

striants

C.

C.

C.

C.

Y

Y

A. camp[iratus.vuhsp.

cump(i(alus órispanicus subsp. argyrophyllus

A.

Y

Y

Y

Adenuearpu.s A. complicatus subsp. anisochilu.s

Cyrisus, average

Y

baetirus ('. bnlan.sne suhsp. eurnpaeas

W

Y

C. arboreta subsp.

CvFisus

Ulex, average

Y

Y

Y

Y

Ulex U. erincladus

8d,ino,sparTUm E. barnadesii subsp. dorsisericeum

.Sparlium S. junceum

Relama, average

R.

14..5711.18 (30) 8.10 14.2011.08 (30) 7.60 15.8O10.77 (20) 4.86

(10) 4.13 12.061:2.06 (110) 17.08 18.3711.00 (30) 5.44 22.0312.32 (6(1) 10.53 16.2014.75 (250) 29.32

15.821069 (30) 4.35 13.4010.81 (10) 6.04 24.1(1.1.0.99

11.2711.07 (30) 9.45 9.45:1:0.67 (30) 7.13 10.3611.27 (60) 12.29

16.0010.67 (10) 4.17

26.0812.36 (30) 9.05

4.3810.39 (50) 8.95 5.3232.15 (60) 40.40

8.50±0.78 (30) 9.14 7.9310.91 (30) 11.44 8.3310.73 (20) 8.77

7.52. 1:0.61 (30) 8.10 6.1010.21 (10) 3.46 8.201.0.68 (10) 8.23 630-1-1.08 (110) 17.14 6.7710.50 (30) 7.45 7.0310.58 (60) 8.26 6.74:1_O.98 (250) 14.55

3.6310.43 (30) 11.95 2.7510.31 (30) 11.20 3.121.0.58 (60) 18.21

4.7510.43 (10) 8.95

(30) 10.58

9.37+0.99

1.29+0.15 (50) 11.29 1 .3410.18 (60) 13.48

5.151O.65 (30) 12.53 5.17+.0.67 (30) 13.04 6.5310.68 (20) 10.39

(30) 10.05 18.2611.62 (60) 8.86 11.6016.48 (250) 55.84

16.6011.16 (30) 7.00 6.101 0.52 (10) 8.47 19.2511.32 (10) 6.85 4.921.0.75 (110) 15.19 17.031.1.71

6.4710.47 (30) 7.30 5.361.0.45 (30) 8.46 5.9110.72 (60) 12.23

9.0510.60 (10) 6.62

13.23+1.53 (30) 11.56

3.01+0.28 (50) 9.38 3.37+0.87 (60) 25.69

C

SB

SB

SB

SI

C

SC

SC

SI

SA

SA

SC

SA

SC

GA

GA

GA

"1'

T

T

R

T

9.673-1.50 (15) 15.48 10.411.12 (15) 10.78 6.0011.05 (10) 1757

(120) 20.70

(25) 10.89 9.9912.07

10.80.1:1.57 (15) 14.51 7.6010.55 (5) 7.21 12.2012.78 (5) 22.75 9.51.1.1.48 (55) 15.53 13.2011.47 (15) 11.16 8.681 .0.95

6.1311.25 (15) 20.31 5.8010.86 (15) 14.86 5.9711.07 (30) 17.87

4.801.0.45 (5) 9.32

17.47+1.36 (15) 7.76

5.12+1.13 (25) 22.07 5.0011.11 (30) 22.28

1W-D-O uW -131. IW-D-O

uW-B-L

I W-D-O

uW -B-L.

I W-D

uW-B-I,

I W-D

uW-B- I.

I W- D

uW-B- 1,

I W-D

uW-13- 1.

I W-D

uW-B-I.

I W- D

uW -B-L

I W-D

uW-B-L.

I W-D

uW-B-L

I W-D

uW-B-1,

IW-D

uW-B- 1.

I W-D

uW-B-L

0

299.84.1_65. 16 (28) 21.73 292.72149.63 (21) 16.96 373.63 6) 29.62

(238) 64.66

454. (30) 13.53 154.86.1 32.37 (II!) 20.90 60439196.71 (IO) 16.00 276.8211 (110) 39.88 600.I0.1.100.22 (30) 16.70 992.51:1271.97 (48) 27.40 492.9

150.521.42.69 (30) 28.36 179.1515636 (60) 31.46

2O7.78154.26 (30) 26.1 I

304.52157.64 (10) 18.93

1823.44+317.89 (30) 17.43

34.12+6.10 (50) 17.89 38.74+12.57 (60) 32.44

(21) 23.06 92.84119.70 (16) 21.22

(28) 37.97 51.68111.92

62. 13123.59

285.46 1 42.34 (30) 14.83 48.06 I.8.48 (10) 17.65 450.00 36.17 (10) 8.04 86.53133.4.2 (110) 38,62 356.91 J_I 13.59 (30) 31.82 604.691 196.76 (48) 32.54 263.851 229.66 (238) 87.115

(30) 25.87 44.17112.77 (30) 28.92 62.43 1:25.17 (60) 4I1.32

80.69120.88

177.69141.28 (10) 23.23

619.571112.56 (30) 18.17

12.31+2.00 (50) 16.26 13.57+3.47 (60) 25.60

(16) 25.24

0.2610.07

(21) 13.60

(28) 29.20 O.18+0.02

0.20+0.06

11.63 1 0.118 (30) 12.71 0.32 1(l.(17 (10) 23.02 0.761 0.13 (I()) 16.69 0.32.1.0.08 (11(1) 24.17 0.6(11_0.16 (30) 26.92 (1.611(1.13 (48) 22.00 0.47.1. O. I 9 (238) 39.63

(1.4(110.10 (30) 24.52 0.3010.05 (30) 18.14 0.35 1D.09 (60) 27.04

0.5810.05 ( I 0) 8.44

0.3410.05 (30) 14.40

0.3810. (50) 29.57 0.3710. (60) 28.92

7.44+1.37 (75) 18.37 7.63+1.42 (90) 18.62

(21) 17.29 2332.301619.00 (16) 26.54

(28) 2134 1722.03_1_297.67

1809.82.1422.46

3698.25 1465.23 (30) 12.58 1014.57.1 183.43 (10) 18.118 5271.94I 557.17 (10) 10.57 1816.761694.83 (11(1) 38.25 4785.05.1951.65 (30) 19.89 7986.00 .22 I 4.08 (48) 27.73 3783.76 1.2694.74 (238) 71.22

1442.341343.95 (30) 23.85 973.411267.62 (30) 27.49 1207.88 1386.34 (60) 31.99

2411.031484.68 (10) 20. I O

(45) 17.19 6.12.1.1.22 (45) 19.91 7.9910.95 (30) 11.95

6.97:1 1.20

8.15 1 (1.9(1 (45) 11.04 5.391.0.8 I (45) 15.09 13.03 1(1.89 (15) 6.79 4.241.1.25 (105) 29.55 12.0712.64 (45) 21.90 I1 901 3.17 (75) 26.65 8.14.1.403 (330) 49.48

6.841 0.60 (45) 8.73 I 1.061.4.58 (90) 41.43

15.271.2.39 (45) 15.62

9.6510.95 (15) 9.80

12215.06L2011.66 47.86:L13.84 (30) 16.47 (45) 28.92

232.15+32-85 (50) 14.15 261.55+75.64 (60) 28.92

(30) 9.65

V-P

L.

Tribe Ceniaveae

Imp/mu, average

nrirrandms

5.67.1.0.55

14.90.10.65 (30) 4.35

Y

7.. lutina

5.501.2.28 (912) 41.45

13.81 14.79

(140) 12.61

(912) 34.68

5.7610.73

143911.19

(10) 9.58

(10) 3.66

(140) 8.29

5.50.1.0.53

14.10 .1 .0.52

(30) 13.92

(30) 4.57 (30) 8.44

5.311.0.74

13.601.0.62

(30) 6.66

(30) 9.81

5.85.1.0.49

5.78 I (L57

13.22 1.0.69

(10) 9.35

(3(1) 5.23

7.15.10.67

1 6.15 1.0.34

(10) 6.60

(I()) 2.09

5.3010.35

II.50.1.0.53

(130) 1 4.18

(10) 4.58

8.12

15.7012.2I

(30) 1(1.53

(130) 1 4.10

6.8510.72

15.5211 .38

(20) 7.19

(30) 8.90

9.551069

19.831.1.52 (20) 7.64

mm

side

Ovary

15.34.1.1.1)2

V-P

Y

13-W

B-W

Y

Y

Y

nun

colour

Flower size

Flower

hispunirus var l~i< nlur L. hispunirus var hispnniras

un,gns7ifbliu.s

L.

L.

ulbn.r

l,.

[.opinas

ArRprolobinm A. aalnunü

Arlenorrrrpns, average

hispunirus subsp. gredensis A. lelonensis

Faxon

A.

"

Table l (continued)

(912) 53.06

8.3114.41

(140) 9.28

8.01.1 0.74

(10) 6.25

7.95.1.0.50

(30) 6.59

8.6210.57

(30) 7.23

8.2710.60

(30) 6.73

8.14-1-.0.55

(30) 5.38

7.05.10.38

(10) 3.57

7.951028

(I()) 5.14

4.10 1.0.21

(130) 19.78

6.2611.24

(30) 6.59

7.55 1.0.50

(20) 9.2)1

7.3510.68

mm

size

Style

Gynoecium

SB

SI3

SI3

SB

SI3

SI3

SC

SI3

SB

type

Style

Stigma

GC

GC

GC

(IC

GC

GC

DT

GA

GA

type

(472) 54.50

8.0914.41

(70) 21.51

4.531.0.97

(5) 15.97

2.80 4).45

(15) 1 4.75

3.8010.56

(15) 9.86

4.20:1 .0.41

(15) 11.57

5.331.0.62

(15) 13.15

4.871.0.64

(5) 0.00

63101.00

(15) 10.15

6.931(1.7(1

(62) 25.33

8.231.2.08

(12) 8.62

7.0010.60

(10) 8.11

6.503E0.53

number

Ovule

IW-D-O

uW-13-1,

I W-D-O

uW-B-1,

IW-D-O

uW-B-I,

1 W-0-O

uW-B-I,

I W-D-O

uW-B-I,

I W-D-O

uW-13-I.

1 W-1)-0

uW-13-1,

I W-I)-O

uW-13-1.

I W-D-O

uW-13-I,

type

Anther

(881) 96.93

398.061385.84

(138) 36.81

633.981233.37

(10) 9.50

48969 1 46.5(1

(30) 11.81

707.17.1.83.52

(30) 24.58

513.7(11.126.30

(28) 18.41

398.77.1.73.40

(30) 15.98

758.8U.1_121.24.

(I0) 10.45

1203.52.1.I25.76

(10) 29.79

137.18.1.4(1.87

(115) 25.98

318.73.1.82.81

(30) 24.07

328.87179.14

(20) 311.75

31334196.34

( Hun"

size

uW Anther

1826

(881) 132.26

134.77.1 178.25

(138) 39.24

77.23 1 .30.30

(1(1) 16.97

51.73.1.8.78

(30) 13.25

69.20.1.9.

(30) 25.29

72.

(28) 23.74

48.37.1.I I .48

(30) 20.18

119.94124.2O

(10) 16.57

94.58115.67

(10) 39.08

36.72.1 14.35

(1IS) 36.93

67.22 124.82

(30) 38.25

70.92127.13

(20) 32.51

64.64121.0 I

dmnL'

size

I W Anther

Androecium

(881) 51.73

0.33 . 1 0.17

(138) 25.81

0.13.1.11.03

(10) 15.55

0.1 11.0.02

(30) 12.02

O. 10.10.01

(30) 13.65

(1.141 (U)2

(28) 1 8.18

(1.12.1 (1.02

(30) 1 6.91

(1.161(1.03

(10) 1536

0.08.1(1.01

(10) 43.80

(1.28.1 O.12

(115) 33.57

021 1(1.07

(30) 42.93

0.23 1.0.10

(20) 29.07

(1.21 I 0.06

IW/uW

Rv =

.

(881) 1 01.08

2664.13.1.2692.99

(138) 35.76

3556.04.1.1271.46

(10) 9.33

2707. 14 .1 252.54

(30) 11.49

388

(30) 24.40

2929.76.1 714.90

(28) 18.20

2235.68 1 406.99

(30) 15.67

4393.62168838

(10) 1 0.23

6490.47.L664.17

(10) 28.45

869.54 I 247.35

(115) 25.91

1 929.75 1500.05

(30) 22.79

1 998.96.1.455.53

(20) 29.91

1 889.90 1. 565.35

Minn'

bionwss/l'I

Anther

6.38

420

3.38

(1335) 72.13

12.87 .1 9.28

(210) 18.97

20.1413.82

(15) 9.80

14.61) 1 1.43

(45)

22.571 3.02

( 45) 1005

1930.1 1.94

(45) 9.63

2(1.6') 1.1.99

(45)

22.31.1.3. 17

(15) 11.31

1 2.711.1.44.

(15)

23.41 13.84

(195) 27.93

8.23 12.30

(45) 15.27

12.16 1.IA.7

(30) 16.32

11.4711.87

size

Pollen

Floral biology in Genisteae

31

B

A

D

C

E

F

G

Fig. 2. Style and stigma types in Genisteae. Al, BI linear and curved at the base of the style; CI, DI, FI, linear and curved at the apex of the style; El, strongly curved and hollowed style; E3 circinate and hollowed style; GI, strongly curved, nonhollowed style; G3, linear, slightly curved and nonhollowed style; A2, globular, crowned stigma; B2, g lobular acrowned stigma; C2, terminal discoid stigma; D2, discoid introrse stigma; E2, G2, tufted stigma; F2. introrse, rectangular-lateral stigma. A Lupinus hispanicus; B Adenocarpus telonensis; C Genista umbellata; D G. hirsuta, El, Cytisus scoparius; E2, E3 C. striatus; F Spartium junceum; GI Cytisus arboreus; G2, G3 C. multiflorus

J. López & al.:

32

B

Fig. 3. Types of androecium in Genisteae. A, upper whorl with linear-oblong basifixed anthers and lower whorl with orbicular or suborbicular dorsifixed anthers (Lupinus angustifolius). B upper whorl anthers similar to the previous case, and lower whorl with oblong or linear-oblong dorsifixed anthers and triangular acuminate apex (Cytisus grandiflorus)

dorsifixed in the 1W. Depending on their morphological features the following types were recognized: (1) linear or oblong-linear uW anthers, and orbicular or suborbicular 1W anthers (Fig. 3A), and (2) uW anthers as in the previous case, and 1W anthers oblong or linear-oblong with a triangular-acuminate apex (Fig. 3B). In Retama monospertna (L.) Bolss. and Spartium junceum we observed anthers with tuft of hairs. The taxa with largest anthers were Spartium junceum and Cytisus striatus, and those with the smallest were Retama sphaerocarpa and R. monosperma (Table 1). Total anther biomass per flower was correlated directly with flower size (Table 3), thus implying that the larger the flower, the greater the volume occupied by the anthers. Anther size differed depending on the whorl: uW biomass as volume was significantly greater than the 1W biomass (Mann-Whitney test U= 6228.00, P< 0.001) in every taxon of the tribe. The largest value of the 1W/uW ratio (Rv) was found in

Floral biology in Genisteae

33

Cytisus grandiflorus (Rv=0.76), and the smallest ratio in the genus Lupinus (e.g. L. albus L., Rv=0.08; Table 1). The largest pollen grains appeared in Spartium junceum and Argyrolobium ' zanonii, and the smallest in Cytisus multifiorus (L HÉR.) SWEET and C. balansae subsp. europaeus (G. LóPEZ & JARVIS) MuÑoz GARMENDIA (Table 1). Considering the tribe as a whole, mean pollen grain volume was 12.87+9.28 µm 3 ; this variable correlated directly with anther biomass per flower and with flower size (Table 3). The direct correlation between pollen grain volume and style length was highly significant (Table 3). However, pollen grain volume correlated inversely with ovule production, although the relation was not significant according to the sequential Bonferroni test. Floral rewards. Nectar. The monadelphous androecium in the tribe implies the absence of an intrastaminal nectary and nectar secretion. However, Retama sphaerocarpa secretes nectar from an extrastaminal nectary. Nectar production in R. monosperma could not be studied. The volume of nectar in this species was low (mean=0.093 µl) when flower production alone was considered, but high in terms of whole plant production (approx. 17.21 ml), because of the large number of flowers per plant that this species can produce (185 000; range 2000—562 500). The concentration of the extracted nectar was high (53.43 + 9.01%) and sugar content was approx. 0.76 + 0.14 mg/µl. Nevertheless, quantifying the nectar was difficult on windy days, when high concentrations were obtained (approx. 80%). P o 11 e n . Mean pollen production per flower considering the entire tribe was 46 481.17 ± 46 029.03, Cytisus striatus and C. arboreus producing most, and Retama sphaerocarpa and Genista triacanthos BROT. the fewest grains (Table 2). At the tribe level uW produced more than twice as many pollen grains (31 531.80 + 26 950.58) as 1W (14 949.37 + 20 666.86). This difference was statistically significant (Mann-Whitney test U = 5823.00, P< 0.001). Pollen production per flower was directly and significantly correlated with flower size and anther biomass per flower (Table 3). Pollen production also showed a linear correlation with ovule production (F=10.02, P < 0.002; Fig. 4); this implies that the more pollen grains produced by the flowers, the larger the number of ovules. The relationship between pollen number and pollen grain size was inverse, but not significant (r = -0.12, P> 0.05, N = 87). When pollen production was considered as total biomass per flower rather than pollen grains per flower, the highest values appeared in Spartium junceum and Cytisus striatus, and the lowest in Retama sphaerocarpa and R. monosperma. This variable also correlated with the floral size, anther biomass per flower, ovary size, style length, and ovule production (Table 3). The percentage of sterile pollen grains per flower was not correlated with pollen production per flower (r = 0.17, P > 0.05, N = 87); this implies that higher or lower pollen production per flower did not necessarily mean higher or lower sterile pollen production. The highest rate of sterility was observed in Cytisus scoparius (33.24 + 36.44%) and in one of the six populations of C. striatus (pop. 14/94), where the entire pollen production was sterile. This was observed in five individuals, and may have been due to environmental factors related to high temperature and water stress during microsporogenesis.

Vol

3440.00 1 .756.97

(15) 21.18 I320Ú.001 2907.13 (15) 22.02 8933.33 1 2882.62 (15) 3227 1 1 1 ( 1 0 . ( 1 0 1 2895.69 (5) 26.09 8780.001 1040.19

(45)-

0

(451_

0

145 )-

(

(15)-

(25) 17.47

3169.561553.86

(5) 19.96

(35) 18.24

7520.00 1. 1500.67

(345)-

(35) 16.87

13231.20.) 6441.00 (125) 48.68

0

53.43519.01

(5) 56.67

(15) 15.82

(15)-

0.093.10.12

3180.0011801.94

21853.33 13456.23

I)

(100) 12650

(125) 5530

(15) 52.17

(45)-

(25) 24.42

1474.16 l 359.94

5623.10) 3109.48

(15) 2157

11646.6742512.50

(15) 29.96

435323 1 1304.32

934030() 1 487264

0

(I5) 30.71

3223.33 1 989.75

735:3.33 1 1571.56 (15) 21.37

(5) 22.05

(5) 11.85

4260.60 1 1112.65

(15) 36.76

472091) 1 1735.02

(15) 23.74

6180.()II 11467.36

(15) 31.27

7880.001 2464.38

(45)-

0.7610.14

(5) 26.12

22200.001 4701.98

0

( 15)-

374119111 1192.12

(15) 27.55

051--

(15) 31.88

971333 I .2676.32

11

3823.80

IW

Production

(10) 40.69

uW

( mg/pp

16560.00 1 4.287.24

Production

Sugar w.

(I0) 25.89

I

(3(I)--

I

11

(It l )

Nectar

(25) 18.34

4643.721 -851.62

(5) 30.23

107(10.0 13234.97

(125) 47.93

18854.30 1 9036.92

(15) 14.42

335(10.001 4828.93

(15) 40.72

1369333 1 5576.16

(15) 20.25

10576.67 1 2142.17

(5) 10.33

12220.00 .1.1261.75

(5) 25.29

15360.00.1 3884.33

(15) 31.18

13653.33 1 4257.58

(15) 21.69

1938().()0 I4202.58

(15) 21.97

3008().(10.166Ú878

(15) 24.56

13453. 33 1 3304.08

(111) 27.44

20383.80 15592.71

Production/PI

(25) 211.09

6.97114.72

(75) 24.18

23.753 5.74

(15) 15.58

64 - 591 111.06

I1.45 .1909 (5) 8645

(345) 44.74

(4590 1 64.87

(15) 928

154.38 114.48

(45) 15. 58

197.32 130.75

(45) 12.20

65.27 17.97

(15) 15.21

98.05:1.14.92

(15) 8.72

102.45 1 893

(45) 33.55

20028167.19

(45) 19.21

101981 1959

(45)16.13

225.97 13645

(45) 26.89

10128 1 2724

(.8)) 15 .08

(75) 25.28

11.05 12.79

(15) 15.58

2731 1.4.26

(345) 52.03

61.00 1 31.74

(15)938

82281 7.72

(4.5) 14.81

91.28 1 13.52

(45) 16.73

28.7414.81

(15) 15.21

38.421 5.84

(15) 8.72

39321 3.43

(45) 31 - 55

103.99

(45) 19.49

47.831 932

(45) 24.65

8(1.371 19.81

(45) 21.49

3898 1 9.54

(30) 32.49

34.7511129

(dmm')

(dinm r )

151.)212291

1 W Biomass

uW I)recess

(125) 242.44

5 88 1 1426

05)-

0.O01 0.00

(15) 154.13

19.81 1305"_3

(15) 155.62

7.91

(5) 93.43

2.48.1 2.31

(5) 112.86

0.80 1 090

(15) 156.18

7.951 12.43

(15) 126.00

1.01 11.27

(15) 174.74

0.221039

(15) 11730

433 1 5.08

(I(1) 172.18

10.05

,r.

Sterility/PI

Pollen

(75) 24.45

34.801.8.51

(15) 15.58

9I.9111 1432

(345) 45.46

206.00 1 93.65

(15)938

236.65 1 22.19

(45) 13.52

288.60 1 39.01

(45) 10.85

94.011 111.20

(15) 1521

136.47

(15) 8.72

141 77 1 1236

(45) 31.73

304.27196.55

(45) 19.28

149.81 I28.89

(45) 18.15

306.34 1 5561

(.1.5) 26.22

141156 1 3677

(31) 17.95

186.67

(dnuu

Biomass/PI

(5 mu8iJ6mou.c

V, S

15 rrurl(iJ6nrus

V, S

(5 ruu7(i/6oru.s

(25) 34.70

967.241 335.64

(5) 11.18

24113 33 1 268.78

(125) 60,20

2892.28 1 1972.47

(15) 152()

7338.331 1115.07

(15) 51.52 NSF, S

3336.00 11718.68 ( 1, rnul(ījlr,r-u.~

(15) 28.78 NSF, S

2550.56 1 1. uuJri/luru.).

(5) 26.81

3.i55..33 1 899.71

(5) 39.59

2159.3 3 1 854.84

(15) 27.9()

2102231 586813

(15) 21.26

N.SIi,S

('. uurlliJlr,ru,v

NSF, S

(. rr rlri/Innry

NSF., S

( . uur8i/7nrus

1

NSF, S

('. nndriJlnnr.e

2858.02 1 607.57

('. nrl(i/lrrrus NSF, S

1348.3 I 133636 (15)2495

NSIi, S,

1973.81 1 536.82 (15) 2721)

( 1. u))rlriJlruu.r

(1(1) 25.17

1I35.281:285.72

l'/O

NSI1. S

('. nudli/lunr.a

NSF, ,S

type

mechanisms,

Pollnuaīon

IX

I rX

X

X-1"X

IX

PX-X

1

1 X

r

I X

1

1 X

IX

PX

IX

syslcw

13rccding

Table 2. Floral rewards (nectar and pollen), pollen presentation mechanisms and P/O ratios in Genisteae. The quantitative characters show the mean value + standard deviation (first line) and the sample size in parenthesis and the coefficient of variation (second line). Pollen release mechanisms: V valvular, P pump, SE specialized explosive, NSE nonspecialized explosive. Pollen presentation: N noto-sternotribic, S sternotribic; Breeding system : X xenogamy, FX facultative xenogamy

-

4(1373.331 1306898 (15) 28.18

0

(45)-

49566.67+11935.7

0

13423.33 16828.61

53620.1.14 .983.76

0 (15) 27.94

(1211) 58.15

(345)-

(45)-

(120) 78.73

55797.50 I 32444.01

0.00 1 0.00

1(16006712352.04

(15) 50.87

37135.83.1:2923738

(25) 31.56

100068.00 1 31096.38 (25) 31.110

(901-

75228.00+23742.11

(15) 15.61)

39573.33I6174.57

(55) 31.71

12990.91 14118.96

(5) 27.05

41420.001 .11211366

(5) 34.76

13300.0014623.31

(15) 1837

II

.

34752.731 11542(11 (55) 33.21

(10.5)-

(15)-

0

5394(1.0()I15555.8(1 (5) 28.84

11

264()().()01 8706.61 (5) 32.98

(15) 21.61

145)-

(45)-

79(121).(1(11 17075.47

0

0

(30) 29.67

6626(1.001 12171.74

7(188.73121(13.51

16147.60 16152.37 (30) 38.10

0.00 10.00

(15) 40.71)

68811.0(1120(10.31

(90)-

1774(1,0111 8045.92 (15) 43.36

(45)-

1151 15.20

(15) 19.72

0

7297.47111(1928

14555.2111286992

(45)-

0

(5) 20.42

14720.00.1.3005.33

(5) 24.511

2196(1.0(11 53811.34

(15)-

(15) 191)1

34366.67-16533.94

(30) 46.53

(15) 17.18

1750.47-1986.50 (30)56.10

3894.6311812.12

61253.331 10521.06

-

(45)-

-

0

-

60173.33 1. 13172.83

.

(15) 30.98

67043.33120772.39

(120) 64.95

92933.33 16036(1.86

(25) 30.31

175296114 153135.73

(15) 2().85

85946.67.1 17916.87

(55) 31.10

47743.641 14848.27

(5) 25.14

95360.00

(5) 25.84

3971)(1.0111 1(12589(1

(15) 17.21

14528(1.(10124998.113

(30) 33.24

23236.33 1 7723.52

(15) 41.54

246211.0111 1(1226.1(1

(15) 17.66

21852.67F3858_8.5

(5) 11.60

36680.00-1 .4253.47

(15) 16.38

9562(1.01)+15659.37

(30) 47.04

5653.10+2704.36

7.03:1 13.14

(15) 136.03

7.03-19.56

(120) 168.13

9.93116.7(I

(25) 106.66

4.66

(15) 109.61

3324136.44

(55) 7647

8.11.16.2(I

(5) 110.46

9.22

(5) 18.10

0.671 0.12

(15) 2(15.52

5.421 11.14.

(311) 206.80

8.29

(15) 134.57

3.791 5.1I1

(15) 181.63

12.78

(5) 64.02

2.12 1 1.36

(15) 265.95

1.69

(30) 18124

7.721.1399

30.56+16.66

307.33193.09

(4:5) 24.16

369.88 1 89.35

(3110) 76.75

561.84 1 431.18

(75) 27115

1151.31 1311.37

(45) 17.17

54(1.33 . 1 92.77

(105) 5412

15O.841 8164

(15) 6.79

702.66

(15) 7.84

161.41 1 1265

(45) 15'93

646.41) 1 1(12.95

(90) 37,92

171.37

(45) 33.39

121.01 140.41

(45) 18.73

221.71 .141.53

(15) 9.80

211.901.20.77

(45) 25.71

2901.80+746.09

(90) 54.53

3.76i_6.82

65.65.1.18.21

(45) 43.16

90.17 138.91

( 3(10) 83.78

420.541352.34

(75) 29.117

869.73 1 252.83

(45) 1626

4692217629

(1115) 51.19

5799129.68

(15) 6.79

539,56I3066

(15) 7.84

81.32 16.37

(45) 16.93

542.541 91.84

(90) 46.65

79(13136.87

(45) 36.57

46.91 1 17.15

(45) 16.72

111.14

(15) 9.80

I42.1141 13.92

(45) 30.47

1652.71.15(13.57

(90) 49.55

111.04

372.981 111.17

(45) 26.62

46O.1)51 22.46

(3(01) 83.57

957.16 1 799.92

(75) 27.81

202E05 1562.13

(45) 15.43

1(109.55I 155.02

(105) 53.17

208.83

(15) 6.79

1242.22

(15) 7.84

242.721 19.(l2

1451 1033

1188.941 194.14

(90) 40.43

250.3911(1124

(45) 34.21

167.93.1 57.44

(45) 17.83

332.85159.36

115) 9.80

35393 1 34.70

(45) 27.37

4554.511.1 1246.66

(90) 52.94

44.32123.46

S

~

l', S

4rlc,,4,<07„s

S

('..euiuu„

N

5758.9111 935.50

(15) 49.16

7431.26 1 3652.94

(120) 71.80

9621O316908.08

~,

125) _)X4

2(1236.721611_38.25

(15) 21.39 .SI±,

6538.931 1398.55 1 1 . xrr, l nuus

(55) 31.28

5(162.151 1583.42

(5) 8.56

7798.64 1 667.18

(5) 29.41

5277_141 1552.08

I IS) 2123

13781.62 1 3339.43

(30) 31.88

3')36.461125479

(15) 37.111

4196.701 1557.04

(15) 22.63

3676.22 1 831.85

,SI'., N

1

1 . „u,6u'/ln,r,s

V, S

C. srnl „,iu.r

N

(' .......l,l/l n,s

V. S

C. sropui.ius

SE, N

C. ,,,ulllJlnrus

NSF), S

C. mulliJlnn,,v

NSV„

(5) 11.44

7672.00 1 877.91 C. „,,,lliJlr,n,s

V, S

5487.16+852.75 (15) 15-54

S

C. mulli/Lr0s

NS1(,

1206.59+632.07 (30) 52.39

X

X

X

X

X

X

X

X

X

X3EX

X

X

Taxon

-

-

1494937+20666.86 (472) 138.25

-

85 47

(472)

-

-

(70) 45.85

2709.70,1,1242.45

(5) 2738

1764.003.48299

(15) 32.31

2282.47+737.42

(15) 16.24

2087.60+33894

(15) 25.92

1909331494.82

(15) 24.19

4131.20 1 999.13

.

(5) 10.86

4940.00:1536.61

(15) 11.17

3254.73 4 363.49

(62) 43.03

12966.13.1.5579.26

(12) 44.56

12041.67.15365.96

(10) 47.97

1 1375.003.5456.20

(I()) 2228

18520.004 4126.02

(15) 22.18

IW

Product ion

31531.80+26950.58

-

--

(15)-

(70) 41.39

(5) 23.61

0

23004.29+9520.80

2044(1.(10+4825.25

(45)-

-

(15) 20.44

0

(225)-

20013 33.1 4091.26

(45)-

0

17453.33:1.5427.29 (15) 31.01

0

17006.67+318556

(15) 2298

(45)-

(15) 18.73

31760.00.17297.14

0

(45)-

(5) 21.01

0

42920.00190 16.76

(62) 29.62

(195)-

(310-

5(1287.10 1. 14894.49

II

I)

(12) 20.77

(45)-

1 169( (1(1 .1 579.30

48041.67.19977.11

0

(15) 4.95

(1(1) 51.40

(30) ..

(IS)-

40120.0(112(162338

0

II

(10) 21.09

(30)-

uV✓

( ag/131)

59230.0(P.L12488.67

Production

Sugar w.

(15) 24.08

I

0

I

Nectar

(45)-

(PO

Vol

Table 2 (continued)

(472) 99.03

46481.17+46(129.(13

(70) 41.24

25713.99J10604.03

(5) 23.12

22204.(1(1.5133.611

(15) 20.74

22295.80:14625.01

(15) 28.61

19540.91[5589.65

(15) 18.71

18916.00.13539.53

(15) 22.33

35891.201.8013.98

(5) 19.10

47860.00 1.9143.10

(15) 5.35

14950.73 1 8011.24

(62) 29.46

63253.231 . 18634.67

(12) 2396

601183:43+ 14398.16

(10) 43.12

51495.00+22204.37

(10) 19.33

77750.00115025.70

(15) 21.89

Production/PI

(472) 225.24

6.28+14.14

(70) 552.35

2.17+11.97

(5) 165.05

0.62-+1.03

(15) 100.75

0.9111191

(15) 225.76

11.73+1.64

(15) 298.87

0.96+2.88

(15) 153.22

0.22.P.0.34

(5) 211691

21.27 L-44.(12

(15) 114.56

0.19 1 0.22

(62) 187.34

508 .111.01

(12) 16216

3.111 5.114

(II)) 153.62

1 1.26_I - 17.30

(10) 96.72

0.37+0.36

(15) 187.07

%

Sterility/HI

Pol Ica

(1305) 134.77

421.241567.70

(210) 34.38

454.671156.29

(15) 9.80

29851-1-2924

(45) 17.10

452.50+77.37

(45) 9.68

336.61:1.32.59

(45) 19.77

353.62169.90

(45) 11.42

697.77.1..79.68

(15) 11.31

54534+61.70

(15) 16.38

273.84. 144.85

(195) 32.59

408.3(1.1.133.(17

(45) 24.75

252.(191 11366

(3(1) 4098

475.(19- 1194.7(1

(30) 14.06

474.17F66.67

(45) 30.29

(dmm')

uW ltiornasx

14.16

(1305) 181.14

203.081367.84

(211)) 44.15

53.97+23.82

(1305) 147.90

624.31+923.33

(210) 35.28

508.63+179.42

(15) 9.80

324.27 1..31.77

25.76 1 .2.52

51438+91.93

(45) 9.73

376.79.1 36.64

(45) 19.90

393.:31 178.25

(45) 1190

788.34

(15) 11.31

608.11 168.80

(15) 16.38

350.04 1 57.33

(195) 34314

513.31 1 174.73

(45) 28.37

313.1 31 162.17

Lupinus

1', S

Lupinus

P, S

lerpinu,v

P, S

1.npiuu.s

P, S

17pinu.v

I', S

l..pinu,~

I? S

('. n1d4Jlor.

V, S

Arlenorurpmc

1'. S

A4enncnr-pu.r

1', S

(30) 38.68

608.07 1 .235.22

l', S

Adenrreurpnx

lypc

Adenuenrpu

(45) 18.23 (15) 9.8(1

Pollination nicchani sins,

(30) 12.36

621.29

(45) 29.81

(dn.))

Biomass/13

(45) 31.79

51.89+16.49

(45) I 1.79

40.19 .14.74

(45) 22.54

39.691895

(45) 16.69

91).57 1. 15.11

.

(15) 11.31

62.771-7.1O

(15) 16.38

76,2(11 12.48

(195) 4142

105.01 1.45.59

(45) 44.33

69.04.149.86

(30) 30.60

132.98+40.69

(30) 9.62

147.

(45) 27.74

(dmin )

'

1W Biomass

(472) 85.20

5816.27+4955.43

(70) 39.83

5822.57:1:2319.02

(5) 45.14

8400.20.13791.82

(15) 24,94

5979.26:1.1491.1 1

(15) 2991

4674.63 1 1398.29

(15) 21.67

3597.18 1 779.52

(15) 24.32

7461.98 1 1814.79

(5) 19.10

7976671 1523.85

(15) 11.41

2176.1)2 1 248.20

(62) 44.10

8291.02

(12) 22.82

8593.7(11 1961.01

(10) 4744.

8029.64

(10) 23.28

I 3276.98.1 3090.84

(15) 16.24

P/l)

X

X

X

I

I X-X

X

X

I IX

X

X

X

system

Breeding

J.

LóPEZ

37

& al.: Floral biology in Genisteae

Table 3. Correlation matrix for the floral attributes of Genisteae. Pairs of self-correlated variables are not indicated (-). *** P < 0.001; ** P <0.01; * P <0.05; ns not significant. Only those coefficients marked *** and ** are significant according to the sequential Bonfeironi test at alpha <0.05 Flower size 1

Anther biomass 2

Pollen size 3

Ovary size 4

Style size 5

Ovule number 6

Pollen number 7

2

.873***

3

.337**

.416***

4

.833***

.716***

.003 ns

5

.758***

.721***

.608***

.370***

6

.267*

.255*

-.264*

.484***

.036 ns

7

.757***

.752***

-.120 ns

.858***

–

.617***

8

.878***

.909***

–

.766***

.705***

.327**

–

9

.785***

.775***

.017 ns

.761***

.510***

–

–

20

25

5

10

15

Pollen biomass 8

.799***

No. ovules/flower

Fig. 4. Regression line for the relationship between ovule number (X axis) and pollen production (Y axis) per flower in Genisteae

P/O 9

38

J.

LÓPEZ

& al.:

Tripping mechanisms and pollen-ovule ratios. In the taxa we studied, three of the four pollen release mechanisms described for Fabaceae ( DELPINO 1868, LEPPIK 1966) were represented (Table 2). 1. Valvular mechanism. The reproductive column (androecium and gynoecium) emerged abruptly from the keel, such that a small cloud of pollen adhered to the ventral part of the insect. The column returned to its initial position upon release from pressure by the insect. More than one visit per flower was possible. 2. Pump mechanism. The pollen was released onto the keel and then pushed to the apex by the 1W stamens. Pollen grains were extruded in small amounts through an apical pore each time the flower was visited by an insect, and adhered to the insect's abdomen. In this case several visits per flower were possible. In Adenocarpus flowers excessive or repeated pressure against the keel could facilitate their two parts separation, simulating a valvular mechanism. The mechanism was specialized in Lupinus, where the keel ended in a long, more or less cylindrical beak, and less specialized in Adenocarpus, where the keel ended in a very short beak. 3. Explosive mechanism. Two main types were recognized: A) S p e c i a l i z e d: When the insect pressed against the keel, the reproductive column emerged abruptly from inside to contact the dorsal surface of the insect's body. The floral parts did not return to their initial position when pressure was over, so that flowers could be visited only once. B) Nonspecialized: When the insect pressed against the keel and wings, the reproductive column emerged from the keel; the next step in some cases was similar to the valvular mechanism, but with the terminal portion of the style remaining outside the keel (Genista sp.). In other cases the corolla was completely displaced (Genista sp., Spartium L.). The pollen adhered to the ventral surface of the insect. The development of the staminal filaments of both verticils, and consequently the part of the insect's body where pollen is deposited, suggested the following floral models of pollen presentation: 1) Noto-sternotribic. During pollination both the back of the insect (nototribic pollen, with a pollinating function) as well as its ventral part (sternotribic pollen, with a feeding function) were impregnated with pollen. Two types were recognized: Type I (type Cytisus striatus). The sternotribic pollen came from the anthers of the uW stamens (whose filaments bend toward the ventral zone of the ovary), except for the F10 stamen, which remained erect and kept on to the half of the style, pointing dorsally. Nototribic pollen was produced by the 1W stamens. Thus, stamens Al, C4 and C5 bent downward (placing their anthers beside the uW), whereas the E8 and E9 anthers were located near or inside the style cavity, below the stigma. This subtype appeared only in C. striatus (Fig. 5A). Type II (type Cytisus scoparius). Similar to the C. striatus, but all anthers located near the stigma (except Al anther) belonged to 1W. This type was found in C. scoparius, C. grandifiorus and C. arboreus. 2) Sternotribic. During pollination all pollen adhered to the ventral surface of the insect's body. Three variants were recognized:

39

Floral biology in Genisteae

A

B

Fig. 5. Flower types in Genisteae. A Noto-sternotribic flower, type Cytisus striatus. B Sternotribic flower, type Cytisus multiflores

Type I (type Cytisus multiflorus; Fig. 5B). The uW pollen was released onto the keel; the 1W anthers extended past the uW anthers and pushed the pollen to the tip of the keel, undergoing dehiscence in the process. As a result all anthers were arranged around the upper part of the style. This model was observed in the rest of the tribe except in Lupinus and Adenocarpus. In taxa that used a nonspecialized explosive mechanism, the more carinal stamens (F10, E8, E9) were located nearer to the stigma than the rest of the stamens, as was especially evident in Spartium junceum. Type II (type Lupinus). As soon as the uW stamens had released their pollen, the filaments twisted downward, such that the anthers were located at a lower level. The 1W staminal filaments lengthened substantially and pushed the uW pollen mass up to tip of the keel, while releasing their own pollen. The stigma was located at almost the same height as the anthers or even slightly above them, so that it appears surrounded by its own pollen. This mechanism was present in Lupinus species. Type III (type Adenocarpus). Similar to the previous process, except that the uW filaments did not twist downward, and the uW anthers remained slightly below the 1W ones. This mechanism was seen in Adenocarpus species. The P/O ratio (Table 2) correlated with all floral characteristics (Table 3), except pollen grain volume. For the tribe as a whole, the mean value of this parameter was 5816.27 + 4955.43. The highest values were found in Cytisus striatus and C. arboreus, and the lowest in Retama sphaerocarpa and Genista anglica. According to CRUDEN (1977), all the taxa can be considered xenogams or facultative xeno g ams. We found no significant differences of P/O ratio between annual (Lupinus) and perennial taxa (Mann-Whitney Test, U=112.00, P = 0.206), although the ratio was slightly higher in the former (6348.32 ± 1929.15) than in the latter (5590.05 ± 4401.5). We found significant associations between the type of pollen release mechanism (valvular, nonspecialized explosive, specialized explosive and pump) and P/O ratio (Kruskal-Wallis test, H=18.94. P <0.001), the highest values appearing in taxa with a specialized explosive mechanism (12 089.00 + 6283.85). These taxa had

40

J.

LÓPEZ

& al.:

significantly higher P/O ratios than those that used a valvular (3926.31 ± 2501.37; Mann-Whitney test, U=1.00. P = 0.019) or nonspecialized explosive mechanism (3193.64 + 1723.40; Mann-Whitney test, U=1.00, P=0.005), but the ratio did not differ from that found in taxa with a piston type mechanism (7380.04 ± 2535.64; Mann-Whitney test, U=13.00, P = 0.46). However, the P/O ratio in the latter was significantly higher than in taxa with a valvular mechanism (Mann-Whitney test, U=10.00, P = 0.021) or a nonspecialized explosive mechanism (Mann-Whitney test, U=8.00, P < 0.001). No significant differences were found between the latter. The relation between the P/O ratio and type of pollen release mechanism changed when variants of the explosive model were not considered. The largest P/O ratio appeared in taxa with a pump mechanism (7380.04 ± 2535.64); this value was significantly different from the ratio in taxa that used a valvular (3926.31 + 2501.37; Mann-Whitney test, U=46.00, P=0.021) or explosive mechanism (5286.6614975.80; Mann-Whitney test, U= 40.00, P = 0.012). No significant differences in P/O ratio were found among members of either of the last two groups (Mann-Whitney test, U= 46.00, P = 0.726). When we looked at different models of pollen presentation, we found the highest P/O ratios in taxa with noto-sternotribic presentation (12 089.00 + 6283.85), these being also significantly higher than those with a sternotribic one (4875.19 ± 2891.12; Mann-Whitney test, U=14.00, P = 0.014). When we looked for relationships between Rv and the mechanism of pollen release (valvular, specialized explosive, nonspecialized explosive and pump), we found the highest values in taxa with a specialized explosive mechanism (Rv = 0.65 ± 0.07). This value differed from those in taxa with a valvular (Rv= 0.37 + 0.11), nonspecialized explosive (Rv = 0.34 ± 0.07) or pump type of mechanism (Rv = 0.16 ± 0.06). However, the Rv for taxa with a valvular or nonspecialized explosive system did not differ significantly (Mann-Whitney test, U= 41.50, P = 0.826). Discussion We found that many of the floral attributes studied in Genisteae from SW Europe were directly correlated. In general, the species with the largest flowers also possess large male (anthers) and female floral parts (ovary and style). These features may have a pleiotropic origin, and thus may be controlled by the same gene(s) involved in the increase in cell size and number (PRIMACK 1987). In the male organs anther size correlated directly with the number of pollen grains it contained, and with pollen size. However, we found no inverse correlation between pollen production per flower and pollen size, as might have been expected ( VONHOF & HARDER 1995). In Fabaceae, some tribes show a direct correlation between pollen number and volume, others an inverse correlation, and in still others no relationship is found (unpubl. data). Negative trade-offs are known in taxa of this family (VONHOF & HARDER 1995) and in other families (e.g. Solanum L., MIONE & ANDERSON 1992). Positive trade-offs have also been reported occasionally in Fabaceae (e.g. Vicia L., ORTEGA-OLIVENCIA & al. 1997) and in distant groups (e.g. Pyrolaceae, KNUDSEN & OLESEN 1993); neutral relationships have also been

Floral biology in Genisteae

41

reported in several taxa (CRUDEN & MILLER-WARD 1981; Polemoniaceae, PLITMANN & LEVIN 1983).

Ovary size correlated with the number of ovules and with style length. A leguminous dried fruit must increase in size as the number of ovules it contains increases, and in most xenogamous plants this number tends to be high, because of the high rate of abortion ( WIENS 1984; unpubl. data). Concerning the relation between the male and female organs, style length and the pollen size were clearly correlated, as were pollen production and ovule production per flower. Larger pollen grains were associated with longer styles; this appeared to bear out the assumption that the greater the volume of the pollen grain the greater the accumulated reserves, which in turn facilitates the formation of longer pollen tubes. This classical hypothesis (DELPINO in DARWIN 1896) was supported by PLITMANN & LEVIN (1983) in Polemoniaceae, WILLIAMS & RousE (1990) in Rhododendron L. and ORTEGA-OLIVENCIA & al. (1997) in Vicia. However, the opposite relationship was proposed by AMict (1830) and upheld by CRUDEN & LYON (1985), according to whom the pollen tubes are nourished through the transmitting tissue or the stylar channel (KNOx 1984, CRESTI & al. 1992, HERRERO & HORMAZA 1996). Pollen tube growth probably requires, at least initially, the use of resources from the pollen grain itself. As CRUDEN & LYON (1985) indicated, correlations of this type tend to be positive when closely related taxa (same genus or family) are compared, as they reflect a phylogenetic rather than a functional relationship. The close relationship between pollen and ovule production was not surprising. In general, xenogamous angiosperms produce large amounts of pollen per ovule ( ORNDUFF 1969). Previous studies of other leguminous species found the same direct correlation e.g. SMALL (1988) in Medicago, ORTEGA-OLIVENCIA & al. (1997) in Vicia, and GALLARDO & al. 1994 in Astragalus subg. Epiglottis (BUNGE) WILLK. From an evolutionary point of view, the combination of androecium, style and stigma type reflects important adaptations related to pollination. Thus the main selective factor that favors fusion of the stamens could be the improvement of the protection of the ovule against possible predator attacks (POLHILL 1976). Anther dimorphism enhances pollen presentation, especially in pump arrangements (ARROYO 1981), because the smaller, dorsifixed 1W anthers push the pollen off of the uW anthers, whereas the latter, which are basifixed, prevent the pollen from being dragged downward when their filaments are retracted. The tufts of hairs on the anthers of some species (e.g. one single hair in Retama monosperma and one tuft in Spartium junceuin) may play a role in pollen transfer to pollinators, or act as a filter that keeps the pollen grains separated from their own stigmas (SCHRIRE 1989). These hypotheses were not examined in the present study. The uW anthers were always significantly larger than their 1W counterparts; as a result Rv (always smaller than 1) can be considered an indicator of the mechanism of pollen releasing. Taxa with a low Rv (0.08—0.26) always use pump type mechanisms. Taxa with a much larger Rv are likely to use the valvular (0.28— 0.58) or nonspecialized explosive systems (0.24—0.45). The highest Rv values (0.60—0.76) were found in taxa with specialized explosive mechanisms. In relation to style morphology, the flowers in taxa in which this organ was strongly curved apically and thickened distally were characterized by noto-

42

J.

LÓPEZ

& al.:

sternotribic pollen presentation. The distal thickening may have adaptive value against potential breakage, as it is this part of the style that receives the impact of the insect's body. Thus in genus Medicago, whose flowers also use an explosive mechanism of pollen release, the styles are very thick in comparison with other genera of tribe Trifolieae (SMALL & al. 1987). In those taxa with valvular or nonspecialized explosive release mechanisms the style is straight and slightly curved apically, the curvature being slightly more pronounce in the latter. In taxa with pump mechanism flowers, the style is mostly straight except for its basal curvature. The ventral cavity on the style (a "pollen carrier" according to POLHILL 1976), plays an important role in collecting a large part of the pollen grains than will be deployed when the release mechanism is triggered. Of the different types of stigma, globose ones are always associated with taxa that use pump type mechanisms, whereas taxa with a specialized explosive mechanism always had a tufted stigma, as also occurs in some taxa that use the valvular system (Cytisus sp., Retama, etc.). In the rest of the taxa introrse stigmas were the most common type among those with nonspecialized explosive mechanisms. Only in Lupinus was the stigma surrounded by a group of sterile hairs, which may act as a barrier to prevent contact between the pollen mass and the stigma of its own flower (WAINWRIGHT 1978). This function may also be fulfilled by the peripheral papillae (larger than the inner papillae) that appear in the rest of the Genisteae. Members of this tribe are basically polliniferous (POLHILL 1976, ARROYO 1981), nectar production being detected in only one of the genera (Retama, from an extrastaminal nectary). Nectar production has also been reported in Erinacea ADANSON, Laburnum FABR., Petteria C. PRESL and Chamaecytisus (POLHILL 1976, BISBY 1981, POLHILL & al. 1981, WEBB & SHAND 1985). The polliniferous character is compatible with the Mediterranean ecosystem in which Genisteae are abundant, because available water is limited by the climate, a factor that conditions nectar production (HERRERA 1985, PETANIDOU & VOKOU 1990). Pollen production per flower was high in the taxa we studied (Table 2); uW anthers produced the greatest amounts of pollen, and released the pollen earlier. In species with noto-sternotribic flowers, the uW pollen fulfilled basically a feed function, whereas 1W pollen was involved in pollination. The absence of such a functional duality indicates a lower degree of specialization. Pollen biomass per flower (Table 1) differed between annual (499.20±174.52 dmm 3 , N=6) and perennial species (574.55 ± 896.25 dmm 3 , N=28), although the difference was not significant (Mann-Whitney test, U=122.00, P = 0.086). Concerning the breeding systems, the high P/O ratios suggest that xenogamy and facultative xenogamy predominate in the studied taxa. This was also the case in the genus Lupinus, although its herbaceous annual habit could suggest an autogamous breeding system (WIENS 1984). Lupinus does not reveal a correspondence between the P/O ratio and the natural or experimental outcrossing rates, as it is a mostly self-compatible genus with a high degree of autofertility (e.g. L. palaestinus Boiss and L. pilosus L., PAZY 1984; L. bicolor LINDLEY and L. nanus DOUGLAS, WILLIAMS 1987; see also KAROLY 1992). In general the outcrossing rates were very low and depended on the genotype, location, season and density of the pollinator (e.g. L. angustifolius, WILLIAMS 1987). In contrast, levels of autogamy in

Floral biology in Genisteae

43

the perennial L. texensis Hook. were very low (less than 5% in natural populations). This taxon is mostly outcrossing (SCHAAL & LEVERICH 1980, SCHAAL 1989). With regard to the mechanism of pollen release, ARROYO (1981) pointed out that lower P/O ratios would be expected in species that use explosive mechanisms because only a single visit of the pollinator is needed, whereas with the other mechanisms several visits can take place. This hypothesis was confirmed by SMALL (1988) for the tribe Trifolieae. In Gen isteae we found no significant differences between the taxa that use explosive or valvular systems. Nor was the hypothesis supported when the two variants we propose for the explosive mechanism were considered: the largest P/O ratios appeared in taxa that use the specialized type. The absence of significant differences in P/O ratio between the nonspecialized explosive and the valvular mechanisms supports the idea that the former system represents an intermediate position between the valvular and the specialized explosive systems. In addition to this, no differences between their Rv values have been found. In conclusion, the highest P/O and Rv ratios were associated with taxa that have noto-sternotribic rather than sternotribic flowers, in contrast with the hypothesis proposed by ARROYO (1981). This research was supported by the Dirección General de Investigación Científica y Técnica (DGICYT PB90—0670) of Spain and Consejería de Educación y Juventud (Junta de Extremadura), Fondo Social Europeo (EIA94—13). We wish to thank A. BOTELLO for the illustrations, an anonymous referee for comments and revision of the manuscript and KAREN SHASHOK for revising the English translation. Appendix 1 Studied material. (Abbreviations: BA Badajoz province, CC Cáceres province, MA Málaga province. The studied attribute for each population is indicated in parenthesis, N nectar, P/O pollen and ovule production and P/O ratio, BP biometry of the pollen, BF biometry of the flower and the gynoecium, BA biometry of the anthers.) Adenocarpus DC. — A. complicatus subsp. anisochilus (Boiss.) FRANCO. CC: Near to Puerto de Berzocana, between Berzocana and Cañamero (334/94; N, P/O, BP, BF, BA). Hoyos (347/94; N, P/O, BP, BE, BA). Zarza la Mayor (313/94; N, P/O, BP, BF, BA). — A. complicatus (L.) GAY subsp. complicatus. CC: Hoyos (348/94; N, P/O, BP, BF, BA). Tornavacas (375/94; N, P/O, BP, BF, BA). Villamiel (346/94; N, P/O, BP, BF, BA). — A. hispanicus subsp. argyrophyllus (RIVAS GODAY) RIVAS GODAY. CC: Guadalupe, near the military base (380/94; N, P/O, BP, BF, BA). Montfragüe, Salto del Gitano (142/94; N, P/O, BP, BE, BA). — A. hispanicus subsp. gredensis RIVAS MARTÍNEZ & BELMONTE. CC: Puerto de Honduras (373/94; N, P/O, BP, BF, BA). San Martín de Trevejo (343/94; N, P/O, BP, BF, BA). — A. telonensis (LoISEL.) DC. CC: Almaraz, Embalse de Arrocampo (258/94; N, P/O, BP, BF, BA). Cañamero, Cerro del Castillo (318/94; N. P/O, BP, BF, BA). Embalse de Valdecañas del Tajo (185/94; N, P/O, BP, BF, BA). Argyrolobium ECKLON & ZEYHER. — A. zanonü (TLTZRA) P. W. BALL. CC: Almaraz, Embalse de Arrocampo (257/94; N, P/O, BP, BF, BA). Cytisus L.— C. arboreus subsp. baeticus ( WEBB) MAIRE. BA: Between Jerez de los Caballeros and Oliva de la Frontera, near to the cross to Higuera de Vargas (33/94; N, P/O, BP, BF, BA). Oliva de la Frontera, Cortijo Los Galvaríes (9/94; N. P/O, BP, BF, BA). Valle de Matamoros, Cerro de San José (26/94; N, P/O, BP, BF, BA). — C. balansae subsp.

44

J. LÓPEZ & al.:

europaeus (G. LÓPEZ & JARVIS) Muy oz GARMENDIA. CC: 2 km from Piornal, in front of a pine-tree plantation (272/94; N. BP). Puerto de Honduras (372/94; N, BP); ibidem (9/95; N, P/O, BP, BF, BA). – C. grandiflorus (BROT.) DC. CC: Valverde del Fresno, Sierra de la ' Malvana (8/95; N, P/O, BP, BF, BA). – C. multiflorus (L HÉR.) SWEET. BA: Alburquerque, Finca el Hito (12/94; N, P/O, BP, BF, BA); ibidem (3/95; N, P/O, BP, BF, BA); ibidem (5/ 95; N, P/O, BP, BF, BA); ibidem (48/96; N, P/O, BP, BF, BA); ibidem (49/96; N, P/O, BP, BF, BA). CC: Brozas, road to Aliseda (15/94; N, P/O, BP, BF, BA). 7 km from Garrovillas, road to Navas del Madroño (16/94; N, P/O, BP, BF, BA). – C. scoparius (L.) LINK. BA: La Bazana (10/94; N, P/O, BP, BF, BA). Castuera, road to Benquerencia (5/94; N, P/O, BP, BF, BA). CC: Ribera de Araya, near to Garrovillas (18/94; N. P/O, BP, BF, BA). – C. striatus (HILL) ROTHM. BA: Alburquerque, Finca el Hito (13/94; N, P/O, BP, BF, BA); ibidem (6/95; N, P/O, BP, BF, BA); ibidem (50/96; N, P/O, BP, BF, BA). Road Alburquerque-Aliseda, near Santuario de Nuestra Señora del Campo (14/94; N, P/O). Alburquerque, road to Valencia de Alcántara (67/94; N, P/O, BP, BF, BA). CC: 7 km from Garrovillas, road to Navas del Madroño (17/94; N, P/O, BP, BF, BA). Echinospartum (SPACH) ROTHM. – E. barnadesii subsp. dorsisericeum G. LóPEz. CC: Guadalupe, near the military base (379/94; N, P/O, BP, BF, BA). Genista L. – G. anglica L. CC: near Piornal (274/94; N, P/O, BP, BF, BA); ibidem, 2 km to Valle del Jerte (275/94; N, P/O, BP, BF, BA). – G. cinerascens LANGE. CC: La Garganta, Embalse de las Angosteras (367/94; N, P/O, BP, BF, BA). Guadalupe, near the military base (331/94; N, P/O, BP, BF, BA). Puerto de Honduras (371/94; N, P/O, BP, BF, BA). – G. falcata BROT. BA: Río Alconero, between San Vicente de Alcántara and La Codosera (78/ 94; N, P/O, BP, BE, BA). CC: Valencia de Alcántara, cross to the road to Jola (69/94; N, P/ O, BP, BF, BA). El Pino de Valencia de Alcántara, on the border with Portugal (70/94; N, P/O, BP, BF, BA). – G. florida L. CC: Road Berzocana – Cañamero, near Puerto de Berzocana (314/94; N, P/O, BP, BF, BA). Guadalupe, near the military base (330/94; N, P/ O, BP, BF, BA). San Martín de Trevejo, chestnut grove (341/94; N, P/O, BP, BE, BA). – G. hirsuta VAHL. BA: Alconera, Sierra de Alconera (92/94; N, P/O, BP, BF, BA). Feria, road to Feria from the main road Badajoz-Zafra (39/94; N, P/O, BP, BF, BA). CC: Cañaveral, between Cañaveral and Arco (132/94; N, P/O, BP, BE, BA). – G. polyanthos subsp. hystrix (LANGE) FRANCO. CC: Montehermoso (279/94; N, P/O, BP, BE, BA). – G. tournefortii SPACH. CC: Zorita, road to Berzocana (309/94; N, P/O, BP, BF, BA). – G. triacanthos BROT. BA: Alburquerque, Sierra del Aguila (129/94; N, P/O, BP, BE, BA). La Codosera, near Convento de las Rocitas (79/94; N, P/O, BP, BF, BA). CC: Puerto de los Castaños (283/94; N, P/O, BP, BF, BA). – G. tridentata L. CC: El Pino de Valencia de Alcántara, on the border of Portugal (71/94; N, P/O, BP, BF, BA). Puerto de los Castaños (282/94; N, P/O, BP, BF, BA). Puerto de Torre de Don Miguel (404/94; N. P/O, BP, BF, BA). – G. umbellata (L'HÉR.) POTRET. BA: Alburquerque, Arroyo Los Ruices (286/94; N, P/O, BP, BF, BA). Retama RAFIN. – R. monosperma (L.) Bolss. MA: Main road Sevilla-Granada, near to Archidona (164/94; N, P/O, BP, BF, BA). – R. sphaerocarpa (L.) Bolss. BA: Aldea de Pallarés, Venta el Culebrín (306/94; N, P/O, BP, BF, BA). Guadajira, Finca La Orden (10/ 95; N, P/O, BP, BF, BA); ibidem (222/96; N, P/O, BP, BF, BA). Trasierra, road to Llerena (298/94; N, P/O, BP, BF, BA). CC: Alía, road to Guadalupe (327/94; N, P/O, BP, BF, BA). Spartium L. – S. junceum L. BA: Aldea de Pallarés (305/94; N, P/O, BP, BF, BA). Badajoz, road to Cáceres (203/94; N. P/O, BP, BF, BA). CC: Almaraz, Embalse de Arrocampo (259/94; N, P/O, BP, BF, BA). Ulex L. – U. eriocladus C. VICIOSO. BA: Between Ermita de Bótoa and Base General Menacho (3/94; N. P/O, BP, BF, BA). Coto Castillo de Miraflores, road between Olivenza and Villanueva del Fresno (8/94: N, P/O, BP, BF, BA). CC: Cáceres, road to Badajoz, km 15 (20/94; N, P/O, BP, BF, BA). – U. minor ROTH. CC: Valencia de Alcántara, cross to Jola

Floral biology in Genisteae

45

(410/94; N, P/O, BP. BF, BA); ibidem, 5 km from the cross to the road of Valencia de Alcántara (411/94; N, P/O, BP, BF, BA); ibidem, on the border of Portugal (409/94; N, P/O, BP, BF, BA). Lupinus L. L. albus L. BA: Bienvenida, road to Fuente de Cantos (160/94; N). San Vicente de Alcántara (77/94; N, P/O, BP. BF, BA). – L. angustifolius L. BA: Jerez de los Caballeros, near Los Templarios hotel (28/94; N, P/O, BP, BF, BA). Between Valle de Santa Ana and Salvatierra de los Barros (34/94; N, P/O, BP, BF. BA). CC: Main Road of Extremadura, Km 280, near the river Búrdalo (41/94; N. P/O, BP, BF, BA). – L. hispanicus var. bicolor MERINO. CC: Main road of Extremadura, Km 272 (190/94; N, P/O, BP, BF, BA). Saucedilla, to Serrejón (173/94; N, P/O, BP, BF, BA). Serrejón-Casatejada road (270/ 94; N, P/O, BP, BF, BA). – L. hispanicus Bolss. & RECTER var. hispanicus. BA: Valle de Matamoros, Cerro de San José (195/94; N, P/O, BP, BF, BA). Between Valle de Santa Ana and Salvatierra de los Barros (196/94; N, P/O, BP, BF, BA); ibidem, 3 km from Valle de Santa Ana (197/94; N, P/O, BP, BF, BA). – L. luteus L. BA: San Vicente de Alcántara (68/ 94; N, P/O, BP, BF, BA). Torrefresneda (66/94; N, P/O, BP, BF, BA). CC: Valencia de Alcántara (76/94; N, P/O, BP, BF, BA). – L. micranthus Guss. BA: cross of the roads Alburquerque-La Codosera (191/94; N, P/O, BP, BF, BA). References AMICT, J. B., 1830: Note sur le mode d'action du pollen sur le stigmate; extrait d ' une lettre de M. AMICT à M. MIRBEL. – Ann. Sci. Nat. 21: 329–332. ARROYO, M. T. K., 1981: Breeding systems and pollination biology in Leguminosae. – In POLHILL, R. M., RAVEN, P. H., (Eds): Advances in legume systematics, 2, pp. 723–769. – Richmond: Royal Botanic Gardens, Kew. BISBY F. A., 1981: Tribe 32. Genisteae (ADAMS.) BENTH. (1865). – In POLHILL, R. M., RAVEN, P. H., (Eds): Advances in legume systematics, 1, pp. 409–425. – Richmond: Royal Botanic Gardens, Kew. CRESTI, M., BLACKMORE, S., VAN WENT, J. L., 1992: Atlas of sexual reproduction in flowering plants. – Berlin, Heidelberg, New York: Springer. CRUDEN, R. W., 1977: Pollen-ovule ratios: a conservative indicator of breeding systems in flowering plants. – Evolution 31: 32–46. - LYON, D., 1985: Correlations among stigma depth, style length, and pollen grain size: do they reflect function or phylogeny? – Bot. Gaz. 146: 143–149. - MILLER-WARD, S., 1981: Pollen-ovule ratio, pollen size, and the ratio of stigmatic area to the pollen-bearing area of the pollinator: a hypothesis. – Evolution 35: 964– 974. DARWIN, C., 1896: The different forms of flowers on plants of the same species. – New York: Appleton. DELPINO, F., 1868–1873: Ulteriore osservazione e considerazioni sulla dicogamia nel regno vegetale. – Pt. I, 1868, 1869. Pt. II, fasc. 1, 1870; Milano. – Atti Soc. Ital. Sci. Nat., 11, 1868: 265–332; 12, 1869: 21–141, 179–233; 16, 1873: 151–349. DEVESA, J. A., 1995: Vegetación y flora de Extremadura. – Badajoz: Universitas. ERDTMAN, G., 1960: The acetolysis method – a revised description. – Svensk Bot. Tidskr. 54: 561–564. FAEGRI, K., VAN dER PUL, L., 1979: The principles of pollination ecology. – Oxford: Pergamon Press. GALLARDO, R., DOMÍNGUEZ, E., MUÑoz, J. M., 1994: Pollen-ovule ratio, pollen size, and breeding system in Astragalus subgenus Epiglottis (Fabaceae): a pollen and seed allocation approach. – Amer. J. Bot. 81: 1611–1619.

46

J. LÓPEZ & al.:

GoRI, D. F., 1989: Floral color change in Lupinus argenteus (Fabaceae): why should plants advertise the locations of unrewarding flowers to pollinators? — Evolution 43: 870—881. PDF HERRERA, J., 1985: Nectar secretion patterns in Southern Spanish Mediterranean scrublands. — Israel J. Bot. 34: 47–58. 1987: Flower and fruit biology in Southern Spanish Mediterranean scrublands. — Ann. PDF Missouri Bot. Gard. 74: 69—78. HERRERO, M., HORMAZA, J. I., 1996: Pistil strategies controlling pollen tube growth. — Sexual Pl. Reprod. 9: 343—347. HIDEUx, M., 1972: Techniques d'étude du pollen au MEB: effects comparés des differents traitements physicochimiques. — Micron 3: 1—31. HoRovITz, A., HARDING, J., 1972: Genetics of Lupinus. V. Intraspecific variability for reproductive traits in Lupinus nanus. — Bot. Gaz. 133: 155—165. KAROLY, K., 1992: Pollinator limitation in the facultatively autogamous annual, Lupinus nanus (Leguminosae). — Amer. J. Bot. 79: 49—56. KAY, Q. O. N., 1987: Ultraviolet patterning and ultraviolet-absorbing pigments in flowers of the Leguminosae. — In STIRTON. C. H., (Ed.): Advances in legume systematics, 3, pp. 317—354. — Richmond: Royal Botanic Gardens, Kew. KNox, R. B., 1984: Pollen-pistil interactions. — In LINSKENS, H. F., HESLOP-HARRISON, J., (Eds): Cellular interactions. Encyclopedia of Plant Physiology, 17, pp. 508—608. — Berlin, Heidelberg, New York: Springer. KNUDSEN, J. T., OLESEN, J. M., 1993: Buzz-pollination and patterns in sexual traits in north European Pyrolaceae. — Amer. J. Bot. 80: 900—913. LEPPIK, E. E., 1966: Floral evolution and pollination in the Leguminosae. — Ann. Bot. Fenn. 3: 299–308. MEEUSE, B. J. D., 1961: The story of pollination. — New York: Ronald Press. MIONE, T., ANDERSON, G. J., 1992: Pollen-ovule ratios and breeding system evolution in Solanum section Basarthrum (Solanaceae). — Amer. J. Bot. 79: 279—287. MOLLER, H., 1883: The fertilisation of flowers. — London: MacMillan. ORNDUFF, R., 1969: Reproductive biology in relation to systematics. — Taxon 18: 121–133. ORTEGA-OLIVENCIA, A., RAMOS, S.. RODRÍGUEZ, T., DEVESA, J., 1997: Floral biometry, floral rewards and pollen-ovule ratios in some Vicia from Extremadura, Spain. — Edinburgh J. Bot. 54: 39–53. PAZY, B., 1984: Insect induced self-pollination. – Pl. Syst. Evol. 144: 315–320. PETANIDOU, T., VOKOU, D., 1990: Pollination and pollen energetics in Mediterranean ecosystems. — Amer. J. Bot. 77: 986—992. PLITMANN, U., LEVIN, D. A., 1983: Character interrelationships between reproductive organs in Polemoniaceae. — Israel J. Bot. 32: 40-41. POLHILL, R. M., 1976: Genisteae (ADAMS.) BENTH. and related tribes (Leguminosae). — Bot. Syst. 1: 143–368. — RAVEN, P. H., STIRTON, C. H., 1981: Evolution and systematics of the Leguminosae. — In POLHILL, R. M., RAVEN, P. H., (Eds): Advances in legume systematics, 1, pp. 1—26. — Richmond: Royal Botanic Gardens, Kew. PRIMACK, R. B., 1987: Relationships among flowers, fruits, and seeds. — Annual Rev. Ecol. Syst. 18: 409–430. PROCTOR, M., YEO, P., LACK, A., 1996: The natural history of pollination. — London: Harper Collins. RICE, W. R., 1989: Analyzing tables of statistical tests. — Evolution 43: 223—225.

Floral biology in Genisteae

47

SCHAAL, B. A., 1989: The population biology of an annual Texas lupine. – In STIRTON, C. H., ZARUCCHI, J. L., (Eds): Advances in legume biology. – Monogr. Syst. Bot. Missouri Bot. Gard. 29: 283-292. – LEVERICH, W. J., 1980: Pollination and banner markings in Lupinus texensis (Leguminosae). – SouthW. Naturalist 25: 280–282. SCHRIRE, B. D., 1989: A multidisciplinary approach to pollination biology in the Leguminosae. – In STIRTON, C. H.. ZARUCCHI, J. L., (Eds): Advances in legume biology. – Monogr. Syst. Bot. Missouri Bot. Gard. 29: 183–242. SMALL, E., 1988: Pollen-ovule patterns in tribe Trifolieae (Leguminosae). – Pl. Syst. Evol. 160: 195–205. – LASSEN, P., BROOKES, B. S., 1987: An expanded circumscription of Medicago (Leguminosae, Trifolieae) based on explosive flower tripping. – Willdenowia 16: 415-437. SOKAL, R. R., ROHLF, F. J., 1979: Biometría. – Madrid: Blume. STIRTON, C. H., 1981: Petal sculpturing in Papilionoid legumes. – In POLHILL, R. M., RAVEN, P. H., (Eds): Advances in legume systematics, 2, pp. 771–788. – Richmond: Royal Botanic Gardens, Kew. STATGRAPHICS, 1994: Statgraphics Plus for Windows version 1.0. – RocKville, MD: Statistical Graphics Corporation. SYSTAT, 1992: Systat for Windows: Statistics, version 5 edition. – Evanston, IL: Systat, Inc. VONHOF, M. J., HARDER, L. D., 1995: Size-number trade-offs and pollen production by papilionaceous legumes. – Amer. J. Bot. 82: 230–238. WAINWRIGHT, C. M., 1978: The floral biology and pollination ecology of two desert lupines. – Bull. Torrey Bot. Club 105: 24–38. WEBB, C. J., SHAND, J. E., 1985: Reproductive biology of tree lucerne (Chmnaecytisus palmensis, Leguminosae). – New Zealand J. Bot. 23: 597–606. WIENS, D., 1984: Ovule survivorship, brood size, life history, breeding systems, and reproductive success in plants. – Oecologia 64: 47–53. WILLIAMS, I. H., 1987: The pollination of lupins. – Bee World 68: 10-16. WILLIAMS, E. G., RousE, J. L., 1990: Relationships of pollen size, pistil length and pollen tube growth rates in Rhododendron and their influence on hybridization. – Sexual Pl. Reprod. 3: 7–17. Address of the authors: JOSEFA LÓPEZ, TOMÁS RODRÍGUEZ-RIANO, ANA ORTEGA-OLIVENCIA, JUAN ANTONIO DEVESA (correspondence), TRINIDAD Ruiz, Departamento de Biología y Producción Vegetal: Botánica, Facultad de Ciencias. Universidad de Extremadura, E-06071 Badajoz, Spain. Accepted April 9, 1998 by V. H. HEYWOOD