|
|||||||
|
|
|
|||||
|
|
|||||||
International Journal of Plant Sciences, May 1997 v158 n3 p282(10)
A comparative study of the reproductive
biology of two Ajuga species (Lamiaceae) in
the Southwest of the
Author's Abstract: COPYRIGHT 1997 University of
The reproductive biology of two closely related Ajuga
(Lamiaceae) species, Ajuga iva (L.) Schreber and Ajuga chamaepitys (L.)
Schreber,
was compared. Ajuga iva
is a perennial, cleistogamous species that occurs
relatively frequently in the southwest of the
Full Text: COPYRIGHT 1997 University of
Introduction
Lamiaceae constitute a large family, ca. 220 genera and 4000 species, of cosmopolitan distribution that is particularly well represented in temperate, warm regions (Cronquist 1981; Hedge 1992). Their peculiar floral structure calls for intricate pollination mechanisms that reflect a long history of adaptive coevolution between plants and pollinators (Huck 1992). Allogamy is the usual reproductive process.
The genus Ajuga comprises 40-45 species of wide distribution in extratropical regions in both hemispheres (Bentham and Hooker 1876
; Melchior 1964). The
corolla consists of a highly developed trilobate
lower lip, and a usually very short upper lip. Ajuga
iva (L.) Schreber and A. chamaepitys (L.) Schreber
are closely related and are the only two species of the genus in the southwest
Ajuga iva is a
perennial, cleistogamous species that occurs in the
Mediterranean region and Macronesia. It produces
mostly cleistogamous (CL) flowers,
and a few chasmogamous (CH) flowers at the end of the
annual cycle. In some years, no CH flowers are produced. This species is
relatively frequent in the southwest
The reproductive biology of several species of Lamiaceae has been widely studied in many respects; however, little mention is made of the genus Ajuga, which so far has been investigated mainly regarding pollination, reproductive system, and fruit dispersion (Proctor and Yeo 1973; Gill 1980; Luond and Luond 1981; Bouman and Meeuse 1992), but in species other than those dealt with here.
The reproductive biology of A. iva and A. chamaepitys was compared, specifically addressing the following: phenology, structure and floral growth, pollen production and germination, nectar production, P/O ratio, breeding system, and seed dispersal and germination.
Material and methods
During February and March 1994, young specimens of Ajuga chamaepitys from a population growing near
the village of Hinojos, province of Huelva, southwest Spain, and Ajuga
iva plants (both 1 yr and several years old), collected at the same location
and in two other places (Hornachuelos, province of
Cordoba, and Moron de la Frontera, province of
Seville, both in southwest Spain) were transplanted and cultivated in a garden
at the Faculty of Sciences of Cordoba.
temperature and photo-period during the November-July vegetative period were 12.6 [degrees] C and 10.5 h, 9.3 [degrees] C and 9.5 h, 8.6 [degrees] C and 9.5 h, 10.5 [degrees] C and 10 h, 12.4 [degrees] C and 11 h, 14.7 [degrees] C and 12.5 h, 19.4 [degrees] C and 13.5 h, 24.2 [degrees] C and 14.5 h, and 27.4 [degrees] C and 15 h, respectively (data from the National Meteorological Service). In 1995, seeds from these plants were allowed to germinate, and the resulting plants were used in most of the experiments, except those involving floral structure and pollen production, which were studied on flowers collected in natural populations. Phenology, pollination, and seed dispersal observations were carried out in the natural populations as well.
CHARACTERISTICS OF THE FLOWERS AND FRUITS
The calyx, corolla, style, stigma, anther, and pollen grain dimensions were determined in flower samples of both species from the natural populations that were fixed in FAA (a 1:18:1 mixture of 40% formaldehyde, 70% ethanol, and glacial acetic acid). The flowers were collected on different plants. Ajuga iva CL flowers were also studied. Numbers of pollen grains were quantified by using 10 immature flowers of each type, from different plants, with closed anthers; the four anthers from each flower were placed in 0.5 mL of 70% ethanol (to let the pollen grains disperse) and shaken. Then, five aliquots of 10 [[micro]liter] from each suspension were withdrawn with a micropipette and placed into a drop of 1% Cotton Blue in lactophenol, the number of pollen grains being counted at 10 x 10 magnification. The counts obtained for each flower were averaged and multiplied by the dilution factor (50).
The germination capacity of the pollen grains was estimated by placing two newly opened anthers from the same flower into 0.01% boric acid and 20% sucrose. After 4 h, grains with a pollen tube as long as at least half the grain diameter were considered to have germinated. Percentage of germination refers to the total number of pollen grains examined.
Fruits collected in cultivated plants were measured and embryo, pericarp, and total fresh mass of nutlets were recorded. Pericarps were recovered after germination and weighed after heating at 60 [degrees] C for 4 h. The embryo mass was calculated as the difference. The water-imbibing capacity of nutlets (mg water/nutlet) was also calculated (as the difference between the nutlet mass obtained before and after placement for 24 h in petri dishes containing two sheets of filter paper soaked in distilled water).
FLOWER AND FRUIT PRODUCTION
The numbers of flowers and fruits produced were estimated from cultivated plants grown in gardens. For A. iva, production of both CH and CL flowers and fruits was recorded for 1-yr-old plants and older plants.
FLOWER DEVELOPMENT
Flower growth was studied in 35 flowers of each type from at least 15 different individuals. Each flower was marked with a thread of a different color when the calyx was ca. 1 mm long, and was assigned day 1. The calyx length of these flowers was measured daily, and these measurements were used as time references in constructing the growth models for other organs hidden in the flower bud (corolla, ovary) and to establish the time of meiosis and germination of pollen grains. For this purpose, 70 flowers of each type at different growth stages were collected and fixed in FAA, and calyx and corolla lengths were measured, as well as ovary dimensions. The data were used to construct growth graphs reflecting the increase in corolla length and ovary volume. The ovary volume was calculated by approximation to the volume of a cube of similar dimension. The data for A. chamaepitys were obtained in April and those for A. iva (CL and CH) in May and June. In order to determine the time of meiosis, flower buds were fixed in a 3:1 mixture of ethanol and acetic acid and subsequently stained with alcoholic carmine (Snow 1963). The germination of pollen grains was observed under the microscope by suspending stigmas (and, in CL flowers, anthers as well) in a drop of 1% Cotton Blue in lactophenol.
BREEDING SYSTEM
A nylon mesh was used to bag 10 inflorescences from 10 different individuals of each species; the remaining inflorescences (control) were always within sight of potential pollinators. Because bagged flowers of A. iva (CH) produced no fruits, 26 newly opened flowers were hand-pollinated with pollen from the same flower and the flowers were subsequently bagged in order to detect self-incompatibility. The presence of agamospermy was investigated by emasculating flowers immediately before anthesis, followed by bagging. The number of seeds produced in each treatment was counted at the end of fructification.
GERMINATION
Germination tests were carried out during the early months of 1995 by using nutlets collected in the spring of 1994 that had been stored at laboratory
temperature. The A. iva nutlets used in these tests came only from CL flowers as a result of the low number of CH flowers and, consequently, the low number of fruits produced. To compare germination of two species, nutlets were placed in petri dishes of 9 cm diameter containing two sheets of filter paper soaked in distilled water. Experiments were carried out in germination chambers under different conditions: in complete darkness (dishes were wrapped in aluminum foil) or under the light of three fluorescent tubes (3 x 15 W), with a photoperiod of 12 h; six constant temperature regimes (5 [degrees] C, 12 [degrees] C, 19 [degrees] C, 26 [degrees] C, 32 [degrees] C, and 37 [degrees] C) and four variable temperature regimes (15 [degrees]/5 [degrees] C, 20 [degrees]/10 [degrees] C, and 25 [degrees]/15 [degrees] C, light/darkness) were used. Experiments were also conducted on nutlets that had previously been heated at 75 [degrees] C for 4 h or cooled at 5 [degrees] C for 3 wk.
Each test involved three replicates of 25 selected (healthy, ripe) nutlets. Dishes were examined on a daily basis for 40 d. The dishes incubated in complete darkness were examined in a darkroom under dim light filtered through two blue and green cellophane sheets. Nutlet viability was estimated at the end of the germination tests; those nongerminated nutlets possessing a normal embryo were considered viable. The percentage of germination values referred to the total number of viable nutlets.
The results were compared by one-way ANOVA. All germination data were transformed (arcsin [square root]%) before analysis.
Results
The natural population of Ajuga chamaepitys that was studied exhibited two germination periods. Germination occurred after the early autumn rains (October), and the resulting plants flowered in December-February. Most seeds germinated at the end of winter (late February and March) and these plants flowered in April-June. Therefore, in February and March plants bearing ripe fruits coexisted with small seedlings at the beginning of their cycle.
Most seedlings in natural populations of Ajuga iva emerged in February-March and immediately formed CL flowers, when stems were hardly 3-4 cm [TABULAR DATA FOR TABLE 1 OMITTED] high and only two or three pairs of leaves had emerged. These 1-yr-old plants almost always produced CH flowers by the end of their annual cycle, usually in late May, and some extending to July. CL flowers were also formed during this period, so alternate verticils with CL and CH flowers, or even a CL flower and a CH flower on the same verticil were observed. Stems died at the end of the annual vegetative period. The following autumn, buds at the base of dead stems formed new branches that immediately began to produce CL flowers. These plants, more than 1 yr old, formed CH flowers by the end of the annual cycle, and CL flowers were also produced during that period. A given plant produced CH flowers in some or all of its inflorescences. Aerial stems eventually withered and sprouted the next autumn, the cycle being repeated. The production of CH flowers is not a general phenomenon: occasionally, some individual plants did not produce any CH flower for years and behaved as strictly cleistogamous plants.
FLORAL CHARACTERISTICS
AJUGA CHAMAEPITYS flowers have a yellow corolla with purple spots; in A. iva, most CH flowers also have a yellow corolla with purple spots; however, some specimens are pinkish purple. There was no significant difference between the corolla size of A. chamaepitys and those of CH flowers of A. iva from plants older than 1 yr (table 1). However, the corolla of 1-yr-old plants (X = 16.12 [+ or -] 2.95 x 7.86 [+ or -] 1.14, range = 11-20, n = 35) was significantly larger (ANOVA, P [less than] 0.01). There was no other significant difference between the CH flowers produced by plants of A. iva of different ages. In A. chamaepitys the style ends in two filiform Stigmatic branches of roughly the same length. The stigma of A. iva CH flowers differed markedly. The two stigmatic branches differed slightly in size; both are laminar, triangular, and usually purple colored, and the lower branch (0.6-1.1 mm) was longer than the upper one (table 1). Calyx, style, and anther differed significantly in size between A. chamaepitys and A. iva CH flowers, even though their dimensions were very similar (table 1).
There was no significant difference in pollen grain size between A. chamaepitys and A. iva CH flowers. In A. chamaepitys, the number of pollen grains per flower ranges from 1810 to 2930 (X = 2445 [+ or -] 439, n = 10) and that of viable grains from 19.9% to 70.8% (X = 35.9 [+ or -] 12.36, n = 24). In A. iva CH flowers, the number of pollen grains is significantly greater (ANOVA, P [less than] 0.01) than in A. chamaepitys (between 3310 and 6788, with X = 4624 [+ or -] 1294 and n = 16) and the percentage of germination of pollen grains ranges from 19% to 72% (X = 43.8 [+ or -] 12.43 and n = 25) and is not significantly different (P [less than] 0.01) from that of A. chamaepitys.
The calyx of A. iva CL flowers was similar to that of CH flowers (table 1). The lower lip of the corolla was never unfolded, so the corolla was always completely enveloped by the calyx. The lower lip of the corolla, style, and stigma length were significantly smaller in CL than in CH flowers (table 1). The anther size and the number of pollen grains formed were related to the time of flowering. CL flowers produced simultaneously with CH flowers (June-July) had anthers of the same size as CH flowers [TABULAR DATA
FOR TABLE 2 OMITTED] (table 1) and produced a similar (P [less than] 0.01) number of pollen grains per flower (X = 4572 [+ or -] 1058, range = 2900-6066, n = 10). CL flowers formed in February had significantly smaller anthers (X = 0.63 [+ or -] 0.044 x 0.65 [+ or -] 0.047 mm, n = 17) than those formed in June-July (P [less than] 0.01), and produced significantly smaller numbers (P [less than] 0.01) of pollen grains per flower(X = 2056 [+ or -] 641, range = 1160-3060, n = 10).
Pollen grains from CL flowers were always smaller (P [less than] 0.01) and had thinner walls than those from CH flowers (table 1). In the CL flowers, anthers were adhered to the stigma throughout floral growth and pollen
grains germinated within the anthers. The mean values ([+ or -] SD) of the P/O ratio for A. chamaepitys, A. iva
CH, A. iva CL (June-July), and A. iva CL (February) flowers were 611 [+ or -] 104 (n = 10), 1156 [+ or -] 313 (n = 16), 1182 [+ or -] 251 (n = 10), and 514 [+ or -] 152 (n = 10), respectively.
FRUITS
The fruits are very similar in both species and consist of four black nutlets (mericarps) with a markedly reticulate surface. There were no differences in mass, dimensions, morphology, or color between the nutlets produced by CH and CL flowers in A. iva. The total fresh mass and embryo mass of the nutlets were significantly smaller in A. chamaepitys than in A. iva. (table 2). The water-imbibing capacity was significantly smaller in A. iva nutlets (table 2). There were also differences in the proportional distribution of mass of pericarp and embryo [ILLUSTRATION FOR FIGURE 1 OMITTED].
FLOWER GROWTH
In A. iva, meiosis took place on days 5-6 in both CL and CH flowers; anthesis in CH flowers took place around day 17 (usually in early morning) and anthers dehisced immediately after flower opening. In CL flowers, pollen grains germinated within the anthers around day 10, and the ripening of pollen occurred nearly 7 d sooner than in CH flowers [ILLUSTRATION FOR FIGURE 2 OMITTED]. In A. chamaepitys, meiosis occurred on days 6-7 and flower growth was somewhat slower than in A. iva. Anthesis took place around days 25-26 (usually in the early morning) and was immediately followed by anther dehiscence [ILLUSTRATION FOR FIGURE 2 OMITTED]. Anthesis and fertilization occurred later than in A. ira CH flowers.
Ovary growth was also slower in A. chamaepitys; however, the ovary volume at the time of fertilization was similar for both species [ILLUSTRATION FOR FIGURE 3 OMITTED]. There were marked differences in the rate of ovary growth between CL and CH flowers of A. iva. The greater slope of the graph corresponding to ovary growth in CL flowers indicated that these ovaries grew much more rapidly, which, together with an earlier fertilization, explained that many inflorescences exhibited much larger fruits from CL flowers than from older CH flowers; as a result, there was loss of uniformity in fruit growth in these inflorescences.
FLOWER AND FRUIT PRODUCTION
The number of flowers produced from cultivated plants of A. chamaepitys ranged from 50 to 286 (X = 127.9 [+ or -] 44.3, n = 18). Anthesis usually occurred simultaneously in the flowers on the same verticil, which usually numbered two. Flowers were open for 2-3 d, depending on the time of fertilization. At roughly 3-d intervals, flowers on a new verticil in each inflorescence opened, once those on the previous verticil had started to wither; therefore, each inflorescence had a single verticil with open flowers at any time. Most fruits consisted of four developed nutlets - only occasionally did some nutlet develop incompletely. The time elapsed between fertilization and fruit ripening was 30-35 d.
The structure of A. iva inflorescences was similar to that of A. chamaepitys inflorescences. Anthesis in CH flowers usually took place simultaneously on the same verticil, usually containing two flowers open 2-6 d, depending on the time of fertilization. Anthesis at a new verticil occurred every 3-4 d, so inflorescences exhibited one or two verticils with open flowers at any time. One-year-old plants produced larger CH flowers than did older plants and the proportion of CH flowers in the two types of plants was also different. In 1-yr-old cultivated plants, the mean number of total flowers produced per plant was 69.5 (SD = 48.1, range = 18-255, n = 37) and 78.5% of these plants produced CH flowers [TABULAR DATA FOR TABLE 3 OMITTED] (the remaining 21.5% only produced CL flowers). For plants that produced CH flowers, the mean number of CH flowers per plant was 13.8 (SD = 9.04, range = 1-32, n = 29), and accounted for 21% (SD = 11.8, range = 2.5-44, n = 29) of total flower production.
Older cultivated plants, with larger stems, produced more flowers. The mean value of total flowers produced per plant was 404.6 (SD = 227, range = 112-992, n = 24). Only 50% of these plants produced CH flowers. For plants that produced CH flowers, the mean number of CH flowers per plant was 14 (SD = 15.8, range = 1-54, n = 15) and accounted for 3.3% (SD = 3.7, range = 0.001-12, n = 15) of all flowers formed. The CL/CH flower ratio increased with plant age and development.
All A. iva CL flowers produced fruits with four developed nutlets, while many CH flowers, from both cultivated and natural population plants, produced no fruits or fruits with some nutlets incompletely developed. The time elapsed between fertilization and fruit ripening was 25-30 d.
BREEDING SYSTEM
Neither species produced seeds from emasculated and bagged flowers (table 3), which indicates the absence of agamospermy. Seed production by bagged flowers of A. chamaepitys was 100%, indicating the presence of
certain self-pollination and autogamy. Anthesis occurred early in the morning and was almost immediately followed by anther dehiscence and the separation of stigmatic branches, which initially lie at a small angle. The stigmatic branches are slightly ahead of the anthers and there is no contact, in principle, between the two organs. The flowers remain in this situation during the first day and also, rarely, during the second day, while fertilization requires insect visitation. In the absence of pollination, the angle between the stigmatic branches widens during the second (third) day and, simultaneously, the lower stigmatic branch gradually bends until it eventually contacts the anthers of the longer stamen pair, thus causing self-pollination. Therefore, if A. chamaepitys is not pollinated by pollen-bearing insects, the flowers self-pollinate [ILLUSTRATION FOR FIGURE 2 OMITTED]. In the flowers produced at the end of the flowering season (June), self-pollination occurs within a few hours after anthesis; also, these flowers usually are open 1 d.
No self-pollination occurs in CH flowers of A. iva, and no seeds were set in bagged flowers (table 3). However, it is self-compatible as shown by the fact that hand-pollination flowers produced 83.6% of seeds (table 3). Anthesis also occurred in early morning and was followed almost simultaneously by anther opening and the separation of the stigmatic lobes ahead of the anthers. This spatial arrangement of anthers and stigmas was preserved while flowers were open, up to 5 d in the absence of pollination. Unlike A. chamaepitys, CH flowers of A. iva require insect visitation to set seed.
The corolla is attractive and possesses nectar guides in both species; their yellow pollen is also attractive and so is the stigma in A. iva, whether yellow or strikingly purple, like the anthers. The highly developed lower lip of the corolla and a very short upper lip facilitate insect access. Neither species produces nectar, or they produce such small amounts that they merely wet the inner walls of the corolla bottom and are impossible to quantify.
Pollinating insects are bees and some small beetles. According to our field observations, insect visits were infrequent in both species. Occasionally, flowers of both species were visited by ants, even though they seemingly took no part in pollination. The visits are probably accidental and the likely result of the persistent search for nutlets by these insects.
FRUIT DISPERSION
Corollas are marcescent in both species and remain, withered, on the fruits until they ripen. In CL flowers of A. iva, the corolla remains as a small transparent membrane that wraps the fruit but does not protrude from the calyx.
Once the fruit is ripe, it splits into four nutlets or mericarps that possess a small fatty excrescence, an "elaiosome" (analogous to seed elaiosomes proper). The fruits from both species are dispersed by the same mechanism
. The propagation of nutlets is effected successively by two agents (diplocoria). The first stage involves the calyx and corolla. Once nutlets are ripe, the calyx bottom shrinks on withering and the fruit, together with the withered corolla, is expelled reaching the calyx mouth. Once the fruit emerges from the calyx, any plant movement caused by the rain, wind, or animals causes nutlets to fall to the ground, aided by the withered corolla. The second stage involves the action of ants (myrmerocoria); at least three different ant species (Messor barbarus L., Aphaenogaster senilis Mayr and Camponotus sp.), attracted by the "elaiosomes," collect nutlets from the two species in the field, whether from the ground or when they are still on the plant.
[TABULAR DATA FOR TABLE 4 OMITTED]
GERMINATION
All of the selected nutlets of the two species that did not germinate in the tests exhibited well-developed embryos and were thus considered viable. Their viability did not change for at least 2 yr. There were marked differences in germination between nutlets of both species under all germination conditions tested. The germination conditions for A. iva nutlets are much more restrictive. Germination occurred only at 19 [degrees] - 32 [degrees] C in the constant temperature tests, and only at 20 [degrees]/10 [degrees] C and 25 [degrees]/15 [degrees] C in the variable temperature tests. The germination percentages obtained were very low, under 16% ([ILLUSTRATION FOR FIGURE 4 OMITTED]; tables 4 and 5). Ajuga chamaepitys nutlets, however, germinated at all the temperatures tested, except at 5 [degrees] C, and always in high proportions, often close to 100%. In the constant temperature regimes, the highest germination percentages were obtained over the range 19 [degrees] - 32 [degrees] C in the light, and at 32 [degrees] C in the dark. In the variable temperature tests, peak values occurred at 20 [degrees]/10 [degrees] C, both in the light and in the dark ([ILLUSTRATION FOR FIGURE 4 OMITTED]; tables 4 and 5).
Preheating for a short time (4 h at 75 [degrees] C) usually increased germination in A. chamaepitys (tables 4 and 5). Germination in A. iva was, however, slightly affected by the heating pretreatment (tables 4 and 5). Precooling had no appreciable effect on the germination of A. chamaepitys or A. iva nutlets.
Except at high temperatures, the absence of light significantly inhibited germination of nutlets of both species (P [less than] 0.01). The effect of
darkness was generally less marked on preheated nutlets and at high temperatures.
[TABULAR DATA FOR TABLE 5 OMITTED]
Discussion
Xenogamy is deeply rooted in the Lamiaceae. Protandry or gynodioecy, especially frequent in this family (Owens and Ubera-Jimenez 1992), maximizes cross-fertilization. Luond and Luond (1981) report gynodioecy in Ajuga reptans; however, this phenomenon is not present in the species studied here. Both species are self-compatible and homogamous.
Morphologically and structurally, Ajuga iva and Ajuga chamaepitys (CH) flowers are very similar but exhibit marked differences in their reproductive behavior. Ajuga chamaepitys forms one type of flower; if the flowers are not visited by an insect within 1 2 d after anthesis, structural changes lead to self-pollination. A similar self-pollination mechanism was reported by Cantino (1985) for another species in this family, Synandra hispidula. In A. chamaepitys the flower duration decreases as the flowering period progresses; this has been previously noted for other species (Munoz and Devesa 1987; Navarro et al. 1993).
Ajuga iva produces two types of flowers: CL flowers that are obligately selfed and CH flowers that are obligately insect pollinated. There are marked differences between CL and CH flowers: the corolla, stamens, and style are smaller in CL than in CH flowers; the anthers of CL flowers adhered to the stigma throughout floral growth; and pollen grains germinated within the anthers. These differences are consistent with those previously reported for other cleistogamous species (Harlan 1945; Schemske 1978; Lord 1979, 1981; Won Lee et al. 1979; Minter and Lord 1983; Schoen and Lloyd 1984; Jasieniuk and Lechowicz 1987; Ruiz de Clavijo and Jimenez 1993). However, CH flowers are produced in small number relative to CL flowers, and no CH flowers are produced by some individuals in some years. The CL/CH flower ratio increased with plant age and development, and the lack of production of CH flowers was less marked in 1-yr-old plants than in older ones.
Both species exhibit autogamy and allogamy, and in A. chamaepitys both processes can occur in one type of flower while in A. iva, the two processes occur in two types of flowers. Autogamy ensures the formation of seeds at the expense of genetic variability. Most of the fruits produced by A. iva come from CL flowers; all such fruits produce four seeds. The formation of some CH flowers with obligated cross-fertilization ensures some degree of
genetic exchange. However, some CH flowers produce no fruits, or fruits with low numbers of nutlets, which indicates that the flowers are not visited or that visits are insufficient to ensure full seed set. The lack of pollinating insects is a limiting factor for fruit production by CH flowers of A. iva.
In A. chamaepitys, all flowers produce fruits with four seeds. The infrequent insect visits to flowers, together with the short period during which cross-pollination can occur, probably determines that selfing was high. The lack of pollinating insects for this species does not restrict fruit production: a large number of fruits were produced by plants that flowered in mid-winter (December-February), when the activity of pollinating insects is minimal. Because the number of flowers that are open simultaneously in the same plant is small in both species, but particularly in A. iva, the incidence of geitonogamy is also probably low.
Phenomena such as autogamy and cleistogamy that make the plants independent of pollinating vectors can be related to severe environments (Moldenke 1975) and to colonizing ability (Levin 1972; Baker 1974).
The higher P/O ratios of CH flowers of A. iva compared with those of A. chamaepitys were related to the obligate xenogamy of the former species and the facultative xenogamy of the latter (Cruden 1977). The P/O ratio obtained for CL flowers of A. iva depends on the time the flowers are produced. Lord (1980) found CL flowers of Lamium amplexicaule also produce varying numbers of pollen grains depending on the period during which the flowers were formed. This confirms the variability of cleistogamous flowers in this respect, i.e., the occurrence of more than one type of CL flower.
The fruits from the two species are morphologically nearly identical. Ajuga iva nutlets are slightly heavier because they possess a larger embryo mass. In both species, nutlets were dispersed by ants, which is commonplace in Lamiaceae (van der Pijl 1972; Luond and Luond 1981; Bouman and Meeuse 1992). However, the germinative characteristics of the nutlets are rather different; the A. iva nutlets appear to be much more strictly controlled, so the nutlets germinate in very low proportions and under very restricted conditions. Water-imbibing capacity was smaller in A. iva, indicating a more impermeable pericarp that can be related to lower germination capacity. The absence of light significantly reduces germination in nutlets of both species. Luond and Luond (1981) note that Ajuga seeds can only germinate with light, and that the pericarp dictates the light requirements for germination. Our results, however, show that nutlets in both species
germinate in darkness, though in much lower proportions than in light.
Low temperatures inhibit the germination of A. chamaepitys seeds, which accounts for no germination in the field during the winter months. If moisture is not a limiting factor, seeds germinate either in autumn (which may not occur if autumn rains are delayed) or, more important, in late winter (February-March). Because A. chamaepitys is an annual species, the number of individuals in a population depends on the number of seeds formed in the previous year, which, in turn, is dictated by the environmental conditions under which the plants developed. Based on the easy germination of nutlets of this species, most seeds are bound to germinate and probably a few will join the seed bank. Plant populations that produce nondormant seeds are more liable to become extinct than those populations that produce dormant seeds, because a soil seed bank may ensure the persistence of a population should a catastrophe (e.g., human action, a severe drought) arise. However, the two periods of seed formation and germination lessen such a risk.
Ajuga iva nutlets exhibit dormancy and few germinate, so probably most join the soil seed bank. Therefore, despite the large numbers of nutlets formed, few seedlings sprout in the field each year, and only in late winter (February-March). The stony nature of the pericarp and its high impermeability probably ensure a long persistence of the nutlets in the soil seed bank. In perennial plants, this character reduces the negative impact of a year of low seed production. The low germination of A. iva nutlets and its perennial character can explain the stability of its population as regards the number of individuals observed from year to year. While A. iva populations are much more abundant than those of A. chamaepitys, they are also very vulnerable and continue to decline, particularly because they are slower to regenerate. Human activity is leading to decimated populations that are taking shelter in rocky or stony places with highly arid, poor soils. The germination differences may account for the qualitative and quantitative differences observed between populations of the two species.
Acknowledgments
I am grateful to Dr. S. Talavera of the
Literature cited
Baker HG 1974 The evolution of weeds. Annu Rev Ecol Syst 5:1-24.
Bentham G, JD Hooker 1876 Genera Plantarum. Vol 2. Reeve
& Co.,
Bouman F, ADJ Meeuse
1992 Dispersal in Labiatae. Pages 193-202 in RM Harley, T
Reynolds, eds. Advances in Labiatae science.
Cantino PD 1985 Facultative autogamy in Synandra hispidula (Labiatae). Castanea 50:105-111.
Cronquist A 1981 An
integrated system of classification of flowering plants.
Cruden RW 1977 Pollen-ovule ratios: a conservative indicator of breeding systems in flowering plants. Evolution 31:32-46.
Gill LS 1980 Reproductive biology of Canadian Labiatae. Phytologia 47:89-96.
Harlan JR 1945 Cleistogamy and chasmogamy in Bromus carinatus Hook. & Am. Am J Bot 32:66-72.
Hedge IC 1992 A global survey of the biogeography
of the Labiatae. Pages 7-17 in RM Harley, T Reynolds, eds.
Advances in Labiatae science.
Huck RB 1992 Overview of pollination biology in the
Lamiaceae. Pages 167-181 in RM Harley, T Reynolds,
eds. Advances in Labiatae Science.
Jasieniuk M, MJ Lechowicz
1987 Spatial and temporal variation in chasmogamy and
cleistogamy in Oxalis
Levin DA 1972 Competition for pollinator service: a stimulus for the evolution of autogamy. Evolution 26:668-674.
Lord EM 1979 The development of cleistogamous and chasmogamous flowers in Lamium amplexicaule (Labiatae): an example of heteroblastic
inflorescence development. Bot
Gaz 140:39-50.
----- 1980 Intra-inflorescence variability in pollen/ovule ratios in the cleistogamous species Lamium amplexicaule
(Labiatae). Am J Bot
67:529-533.
----- 1981 Cleistogamy: a tool for the study of
floral morpho-genesis, function and evolution. Bot Rev
47:421-449.
Luond B, R Luond 1981
Insect dispesal of pollen and fruits in Ajuga. Candollea
36:167-179.
Melchior H 1964 Syllabus der
Pflanzenfamilien. Band II.
Minter TC, EM Lord 1983 A comparison of cleistogamous and chasmogamous
floral development in Collomia grandiflora Dougl. ex Lindl.
(Polemoniaceae). Am J Bot 70:1499-1508.
Munoz A, JA Devesa 1987 Contribucion
al conocimiento de la biologia
floral
Navarro L, J Guitian, P Guitian 1993 Reproductive biology of Petrocoptis
grandiflora Rothm. (Caryophyllaceae)
a species endemic to northwest
Owens SJ, JL Ubera-Jimenez 1992 Breeding systems
in Labiatae. Pages 257-280 in RM Harley, T Reynolds, eds.
Advances in Labiatae science.
Proctor M, P Yeo 1973 The
pollination of flowers. Collins,
Ruiz de Clavijo E, MJ Jimenez 1993 Cleistogamy and chasmogamy in Ceratocapnos heterocarpa (Fumariaceae). Int J Plant Sci 154:325-333.
Schemske DW 1978 Evolution of reproductive
characteristics in Impatiens (Balsaminaceae): the
significance of cleistogamy and chasmogamy.
Ecology 59:596-613.
Schoen DJ, DG Lloyd 1984 The selection of cleistogamy and heteromorphic diaspores. Biol J Linn Soc
23:303-322.
Snow R 1963 Alcoholic hydrochloric acid-carmine as a stain
for chromosome in squash preparations. Stain Technol
38:9-13.
van der Pijl L 1972 Principles of dispersal in higher plants. Springer,
Won Lee C, HT Erickson, J Janick 1979 Cleistogamy in Salpiglossis sinuata. Am J Bot 66:626-632.