| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Reproductive Biology |
Universidade de Lisboa, Museu Nacional de História Natural, Jardim Botânico, Rua da Escola Politécnica no. 58, 1269–102 Lisboa, Portugal; Area de Biodiversidad y Conservación, Universidad Rey Juan Carlos—ESCET, Tulipán s/n. 28933 Móstoles, Madrid, Spain
Received for publication August 28, 2006. Accepted for publication June 25, 2007.
ABSTRACT
Natural hybrids between rare and common sympatric species are commonly eradicated to avoid the potential extinction of the rare species, although there is currently no clear predictive framework to quantify this risk. As hybrids can have intrinsic value as new evolutionary pathways, further knowledge on the factors controlling hybridization is needed. In this study we evaluated the role of pollination patterns in hybridization events in two sympatric populations of Narcissus cavanillesii and N. serotinus in Portugal. Narcissus cavanillesii is a rare species, while N. serotinus is widely distributed across the Mediterranean. The hybrid, N. xperezlarae, is quite frequent in southeastern Spain but is scarce in Portugal. Reciprocal manual crossings confirmed compatibility between the two species, although hybridization was more successful when N. cavanillesii participated as female. Narcissus cavanillesii and N. serotinus only shared one pollinator, Megachile sp. (Hymenoptera), which had low visitation rates and high flower constancy. No single isolation mechanism was fully effective in preventing hybridization. Temporal displacement of flowering peaks, strong pollinator specificity, and high flower constancy in the shared pollinator all contributed to limiting hybridization in this site. In other sympatric occurrences, different phenological windows and pollination assemblages may allow greater frequency of the hybrid.
Key Words: Amaryllidaceae • breeding system • conservation biology • flowering phenology • hybridization • Narcissus cavanillesii • Narcissus serotinus • pollination ecology
Natural hybridization and introgression are important processes in plant evolution (e.g., Ellstrand, 1992
; Mayr, 1992
; Levin et al., 1996
; Arnold, 1997
), although they can have different outcomes. In some cases, they negatively affect biodiversity through a decline in reproductive fitness (Antilla et al., 1998
) or the genetic assimilation of threatened plant species by congeners (Carney et al., 2000a
). The effects can be particularly deleterious when dealing with rare species with narrow distribution such as island endemics (Levin et al., 1996
; Rhymer and Simberloff, 1996
). Moreover, when these species have weak pre-reproductive barriers or have recently experienced interspecific contact, hybridization can cause extinction in less than five generations, making it one of the fastest processes to threaten the existence of a species (Wolf et al., 2001
; Rieseberg, 2006
). However, not all hybridization events lead to extinction. On the contrary, hybridization may result in the development of new adaptive traits allowing the colonization of new habitats (Johnston et al., 2004
), fitness enhancement (Burke et al., 1998
), or the origin of new hybrid lineages (Grant, 1981
; Arnold, 1997
). Hybridization can also reinforce reproductive barriers through natural selection for conspecific gene flow (Arnold, 1992
), the creation of stable hybrid zones (Barton and Hewitt, 1985
), or the formation of introgressive races (Anderson, 1949
). Most importantly, hybridization does not pose a threat when congeneric species have naturally coexisted for many generations (Rieseberg, 2006
), and in these cases sympatric zones can be important areas for evolutionary processes (Cozzolino et al., 2006
). Thus, our perspective may change depending on whether we consider the short-term (conservation) or the long-term (evolutionary) effects of hybridization.
Hybridization is not the general outcome whenever congeneric species come into contact because there often are pre- and postmating barriers that prevent hybridization (Templeton, 1989
). Nevertheless, certain families and genera seem to be predisposed for the origin of hybrids (Ellstrand et al., 1996
). In the genus Narcissus L. (Amaryllidaceae), high frequencies of natural hybrids have been reported (e.g., Fernandes, 1968
; Mathew, 2002
), and hybridization has been suggested as an explanation for the phenotypic variability within and between populations (Fernandes, 1968
). Within this framework, we can ask: Does hybridization threaten the populations of rare Narcissus species or, on the contrary, does it lead to new lineages with no prejudice to the parental species? What factors limit hybridization in Narcissus? That artificial crosses can easily give rise to hybrids, even in unrelated or morphologically different species (Fernandes, 1968
; Blanchard, 1990
), suggests that premating barriers play an important role in isolating species. Thus, pollinator relationships and floral morphology apparently contributed to the diversification of the breeding system among Narcissus species (Pérez-Barrales et al., 2003
; Graham and Barrett, 2004
; Barrett and Harder, 2005
). However, pollination interactions between co-occurring species are poorly understood (Barrett et al., 1996
, but see Pérez-Barrales et al., 2006
), and to date no study has evaluated their role in maintaining species boundaries in sympatry.
In this paper, we have focused on the hybridization between Narcissus cavanillesii A. Barra & G. López and N. serotinus L for three reasons: First, the hybrid N. xperezlarae Font Quer, of natural origin, has been long reported (Pérez Lara, 1882
; Font Quer, 1927
) and new populations have recently been documented (Marques et al., 2005
). This shows that hybridization between these two species is not a sporadic event but is quite frequent in nature. Second, the hybrid is not present in all parental sympatric populations. When present, its abundance varies considerably, from small, scarce patches in the middle of the parental populations in the southwestern Iberian Peninsula to large areas in the southeastern Iberian Peninsula where the progenitor N. cavanillesii is currently absent (Soler, 1998
; Marques et al., 2005
). The different abundance of the hybrid suggests that premating barriers must act differently in restricting the formation of the hybrid. Third, N. cavanillesii is endemic to the southwestern Iberian Peninsula and northern Africa, while N. serotinus is widespread throughout the Mediterranean basin. In Portugal the interactions between these two species are particularly important because N. cavanillesii is a threatened species with only two populations (Rosselló-Graell et al., 2003
). The hybrid has a low abundance in these two populations (Marques et al., 2005
), and we do not know whether hybridization poses an additional threat to the subsistence of N. cavanillesii.
Our hypothesis is that hybridization between threatened N. cavanillesii and widespread N. serotinus in Portugal is severely limited by differences in pollinator relationships. Thus, the purpose of this study was to identify the factors that limit hybridization between these two species and to assess the relevance of this process in the conservation of the Portuguese populations of threatened N. cavanillesii. Specifically, we aimed to (1) understand flowering phenology patterns in N. cavanillesii and N. serotinus, (2) determine the dependence of fruit and seed set on insect pollination, (3) assess the degree of interspecific compatibility through artificial crossings, and (4) analyze the role of pollinators in natural hybridization.
MATERIALS AND METHODS
Study area
The studied populations are located in Ajuda (Alentejo) near the Guadiana River (38°46' N, 7°10' W, altitude 
183 m) in Portugal. The area has a mean annual rainfall of 581 mm and a mean annual temperature of 16°C (Instituto de Meteorologia, 1961–1990
). Relative air humidity varies from 55% in summer to 82% in winter (Instituto de Meteorologia, 1961–1990
). The area is composed of holm oak woods with Quercus rotundifolia Lam. and Q. suber L. and Mediterranean riparian communities dominated by several forms of highly branched riparian shrubs like Nerium oleander L. and Flueggea tinctoria (L.) G. L. Webster. Narcissus serotinus is more abundant than N. cavanillesii (60 : 20 ratio of reproductive individuals, respectively), but while N. cavanillesii has the capability of clonal reproduction, N. serotinus usually occurs as isolated individuals. The population in the area is affected by the Alqueva Dam, a large water reservoir, but the impacts of the dam on these species are still being evaluated (Rosselló-Graell et al., 2003
).
Plant species
Narcissus cavanillesii and N. serotinus (Amaryllidaceae) are two perennial geophytes. In Portugal, the former is a "critically endangered" species according to the World Conservation Union (IUCN) categories because of its small area of occupancy, small population size, and fragmentation (Rosselló-Graell et al., 2003
), and it is listed under Annexes II and IV of the Habitats Directive of the European Union (EEC 92/43). In contrast, N. serotinus is widely distributed in the southeast of the country. Both species bloom in early autumn and generally produce only one flower per individual. Narcissus cavanillesii has bright yellow flowers, a very short floral tube (less than 2 mm), and an inconspicuous corona that allows an open corolla and total exposure of sexual structures. Narcissus serotinus has white, erect flowers that are highly scented even at night. Flowers have a long, narrow floral tube that hides internal structures and a short (less than 2 mm) yellow corona. In N. cavanillesii, the stigma becomes receptive before the stamens mature (protogynia), whereas in N. serotinus pollen and stigma are functional at the same time (Marques et al., in press
). Both species have actinomorphic flowers with six stamens located in two whorls. The upper stamens are higher than the style or at the same height, but rarely exceed the floral tube. The lower stamens are shorter than the style, and the nectaries are located above them. In both species, flowers last 4–6 d. The fruit is a small ellipsoidal capsule that releases the seeds by three longitudinal splits when the pericarp is dry. In the studied population, N. cavanillesii has 2n = 28 chromosomes, while N. serotinus has 2n = 10 chromosomes (I. Marques, unpublished data). The species belong to different sections that are not phylogenetically closely related (Graham and Barrett, 2004
)—N. cavanillesii belongs to sect. Tapeinanthus (Herbert) Traub. and N. serotinus to sect. Serotini Parl, according to Webb (1980)
.
Narcissus xperezlarae presents two floral morphologies: one is characterized by yellow flowers with a short perianth tube and the other presents pale yellow flowers with a higher perianth tube and long filaments with the stamens slightly above the corona. Flowers are actinomorphic and have a similar duration as those of the parental species. Pollen and stigma are functional at the same time, and although 30–60% of pollen grains present in each anther are sterile, 40–60% of the individuals form viable seeds (I. Marques, unpublished data).
Flowering phenology
Flowering phenology was assessed throughout the flowering period for both species from 15 September to 1 November 2001. Flowers were censused daily in 3356 and 2889 individuals of N. cavanillesii and N. serotinus, respectively. The census covered the whole population of N. cavanillesii but only part of the population of N. serotinus. In the latter, individuals were randomly chosen across the population. Flowering duration, flowering peak, and flowering synchrony were studied in each population. Flowering duration was estimated as the number of days the population remained in bloom. Flowering peak was the date when the maximum number of open flowers was registered. Within-population floral synchrony was calculated as follows (Albert et al., 2001
):
|
| (1) |
|
| (2) |
). Daily weather data on temperature and rainfall from 27 September to 1 November 2001 were obtained from the Meteorological Station of Caia (Portugal), located 9 km northeast of the populations.
Dependence on pollinators and interspecific hybridization
Controlled pollinations were carried out to determine the role of insect activity in the reproduction of each species reproduction and in postmating reproductive barriers to hybridization. The following treatments were used: (1) passive autogamy—flowers were bagged before anthesis; (2) induced autogamy—pollination with pollen from the same flower, followed by bagging; (3) xenogamy—emasculation and pollination with a pollen mixture from 10 plants in the same population, followed by bagging; (4) interspecific xenogamy—emasculation and pollination with a pollen mixture from 10 plants of the other species, followed by bagging; (5) control—nonmanipulated flowers. A total of 60 randomly selected flowers were used in each treatment. All treated flowers except controls were bagged with 1-mm mesh nylon tulle to exclude pollinators. Flowers were monitored for fruit set after anthesis. Mean fruit set, mean seed number per capsule, and mean seed mass were obtained for each treatment. To evaluate seed viability, four replicates of 25 seeds from each pollination treatment were sown in Petri dishes on moistened filter paper. Seeds from the passive and active autogamy treatments were bulked together for this assessment because they both had the same pollen source. Seeds were incubated in growth chambers with a constant temperature of 15°C and 16-h photoperiod. Previous experiments had shown that both parental species have optimal germination rates under these conditions (I. Marques, unpublished results). Every 2 d we counted and removed germinated seeds until no additional seeds germinated. At the end of each assay, we assessed final germination percentage. For the analysis of the response variables, chi-square tests were used when the variables complied with requirements of normality and homocedasticity, whereas nonparametric Mann–Whitney or Kruskal–Wallis tests were used otherwise. Significances of the multiple comparison of means were corrected with the Bonferroni adjustment (Sokal and Rohlf, 1995
). All statistical analyses were carried out using SPSS 11.0 (SPSS, Inc., Chicago, Illinois, USA).
Further studies concerning the fitness of experimental hybrids were not performed because the hybrids require several years to reach the reproductive stage (Blanchard, 1990
).
Insect activity and behavior
In 2001, we observed and captured pollinators of N. cavanillesii and N. serotinus during the co-blooming period. For each species, two 1 x 1 m plots with similar densities were randomly established within the area of occupancy. A total of 301 and 532 flowers of N. cavanillesii and N. serotinus, respectively, were recorded within the plots. Data were gathered in 20-min observation sessions, spread evenly between 11 and 22 October 2001. An average of 24 diurnal and 21 nocturnal observation sessions were carried out each observation day for a total of 45 h of diurnal observation (on 6 d) and 55 h of nocturnal observation (on 8 d) in each species. Pollinator activity was assessed during daylight from 1000 to 1800 hours Greenwich Mean Time (GMT) and after dusk from 1900 to 0200 hours GMT. Pollinator activity was simultaneously monitored in one plot of each species by two different observers. Every 20 min, the observations changed between the two plots of each species, and every 2 h the observers interchanged the monitored species. The same plots were used on all observation days. In each observation session we recorded: (1) insect identity; (2) insect behavior, i.e., whether the insect acted as a legitimate pollinator or a nectar thief and the reward foraged (nectar or pollen); (3) time spent at each flower; (4) number of open flowers visited; and (5) flower constancy estimated as the frequency of visits in which the visitor had previously visited a flower of the same species. Only insect species that visited three or more flowers were included in the analyses. In each observation session, the number of open flowers per plot was recorded to determine the correlation between number of open flowers and number of insect visitors.
RESULTS
Flowering phenology
Autumn 2001 had typical temperatures and rainfall during the blooming period of both species. Daily temperature varied from 16.4°C to 30.6°C, with a mean of 22.1°C. Maximum precipitation fluctuated between 11.9 mm in September and 38.1 mm in October, with a total of 18 rainy days. The flowering period was short in both species. Figure 1 shows the percentage of flowering plants throughout the flowering period in 2001. Narcissus cavanillesii bloomed first and had a continuous flowering season of 32 d from 27 September to 28 October. Its flowering peaked 2 wk after the first day of flowering. Narcissus serotinus started flowering 16 d later than N. cavanillesii and had a flowering season of 21 consecutive days from 12 October to 1 November. Its flowering peaked on 17 October. The pattern of the percentage of flowering plants was asymmetrical and skewed to the right in both species (N. cavanillesii, g1 = 0.47, ts = 5.83, P < 0.001; and N. serotinus, g1 = 1.78, ts = 3.57, P < 0.001). Within-population flowering synchrony was higher in N. serotinus than in N. cavanillesii (S = 0.45 vs. 0.23, respectively). Flowering overlap between N. cavanillesii and N. serotinus was 15 d (Fig. 1), representing 81% of the flowering season of N. serotinus and 53% of the flowering season of N. cavanillesii.
|
|
Results obtained with interspecific crosses differed depending on which species acted as pollen recipient, although in both cases viable seeds were produced. When N. cavanillesii was the pollen recipient, fruit set, seed number, and seed germination percentage were significantly higher, whereas seed mass was significantly lower (Table 2).
|
Insect activity and behavior
The same pollinator assemblage was observed during the 11 d of observation, and the behavior of each pollinator remained invariable throughout this period. The two species of Narcissus were visited by different pollinator species. Open N. cavanillesii flowers were mainly visited by Hymenoptera, like Halictus sp. and Megachile sp. Less frequent visits were made by some Diptera and some honeybees (Table 3). Most pollinators collected nectar, but Syrphidae and Vespidae collected pollen. The number of flowers visited differed among visitors. Halictus sp. visited more flowers and spent on average more time at each flower than the other groups of insects (Table 3). Narcissus serotinus attracted more insect visitors than N. cavanillesii (Table 3). Its main visitor was Meligethes sp. (Coleoptera) followed by Eristalis pratorum (Syrphidae) and Megachile sp. (Hymenoptera). Butterflies, which were not observed in N. cavanillesii, were frequently seen in N. serotinus. However, they were probably opportunistic visitors because they were not observed carrying pollen or touching the stigma. Most visitors collected nectar, but the adults of Meligethes sp. fed on pollen. No signs of nectar robbing were observed in either of the studied species.
|
A significant positive correlation between the number of flowers visited in each census and the number of open flowers was found in N. serotinus but not in N. cavanillesii (Fig. 2). Flower constancy was high because pollinators mostly visited other flowers of the same species (Table 4). The only insect common to both species was Megachile sp. (Table 3). When visiting N. cavanillesii flowers, this pollinator switched allegiance to surrounding individuals of N. serotinus in only 2.5–2.8% of the cases. Similarly, when visiting N. serotinus flowers, it switched allegiance to surrounding individuals of N. cavanillesii in only 3.4–3.9% of the cases. In both scenarios, Megachile sp. switched more often to nearby flowers of Ranunculus peltatus Schrank and Diplotaxis catholica (L.) DC. (Table 4).
|
|
Incidence of and limits to hybridization in threatened Portuguese populations of N. cavanillesii
Although certain genera are more prone to hybridization than others (Ellstrand et al., 1996
), species are usually isolated by a number of barriers to gene flow (Charlesworth and Charlesworth, 2000
). With no geographic or habitat barriers, sympatric populations of congeneric taxa may achieve a high degree of isolation by opening their flowers at different times (Wendt et al., 2002
). However, in the populations of N. cavanillesii and N. serotinus studied here, a substantial overlap in the flowering periods in the monitored year made temporal isolation an ineffective premating barrier (Fig. 1).
Pollinator activity and behavior is another important isolation barrier in flowering plants (Campbell et al., 2002
). Because sympatric populations can be pollinated to a different extent by diverse groups of insects with different body shapes, sizes, and behavior, the incidence of hybridization will depend on the frequency of interspecific gene flow and the effectiveness of pollen transfer (Campbell et al., 2002
).
Our results revealed no differences in fruit set between control and cross-pollinated flowers of parental species, suggesting that these species were not pollen limited in the studied population. In fact, we observed that flowers were often visited more than once in a single observation session (Fig. 2). Pollinators increased reproductive success because fruit set was significantly higher in both species when flowers were insect-pollinated than when flowers were bagged.
In our study, the pollinators of N. cavanillesii were less diverse than those of N. serotinus. The greater diversity of pollinators of N. serotinus could be explained by a greater density and attraction of its flowers and a pollinator community that reacts to their abundance and floral traits (Fig. 2). The flower arrangement in an actinomorphic, erect pattern provides a large platform where insects can land, and the contrasting corona together with the presence of UV patterns assist pollinators in locating the sexual organs (Marques et al., in press).
Species in the same habitat with overlapping flowering periods may evolve floral features attracting different pollinators (Anderson and Schelfhout, 1980
; Grant, 1994
; Wendt et al., 2002
). Despite having flowers with contrasting colors, UV patterns, and shapes, the studied species were not completely ethologically (Grant, 1949
) or mechanically (Stebbins, 1950
) isolated because they shared one pollinator (Megachile sp.). This pollinator had high flower constancy (Table 4) decreasing the chance of interspecific gene flow. Flower constancy to specific species depends on the attractiveness of floral signals and the availability of pollen and nectar, and allows pollinators to minimize energy and time searching for food (Levin and Anderson, 1970
). In our study, the flowering pattern, which was skewed to the right in both species, provided a high concentration of flowers when each species began to flower; this high concentration of flowers can attract pollinators (Thomson, 1980
; Torres et al., 2002
) and maintain insect–flower constancy.
The significant viable progeny obtained in the interspecific crossings suggests that postmating compatibility barriers between the two species, both physiological and genetic (different chromosome number), do not prevent hybridization. Interspecific crossings showed that reproductive barriers were weaker when N. cavanillesii was the pollen recipient because fruit set and number of seeds per fruit were much higher in this case (Table 2). Moreover, hybrid seeds obtained from N. cavanillesii also germinated better (Table 2). Therefore, we can predict that hybrids are more likely to arise from N. cavanillesii mothers and that these hybrids have a greater fitness than those that arise from N. serotinus mothers, at least in early life stages. Studies are currently underway to evaluate the fitness of the hybrids under natural conditions and the role of later-acting postmating barriers in subsequent life stages.
In conclusion, in the population and year studied, no single isolation mechanism was fully effective in preventing hybridization between N. cavanillesii and N. serotinus. However, temporal displacement of flowering peaks, strong specificity of the main pollinators, and, especially, high flower constancy in the only shared pollinator all limited hybridization in this particular study site where natural hybrids were scarce.
Incidence of and limits to hybridization in other years and populations
The results of this study are restricted to one year and one location, and the potential factors limiting hybridization probably act differently depending on the prevailing environmental conditions or population characteristics in each year and/or location. Because species isolation is mainly determined by insect specificity, we expect between-year variation in hybridization events to be largely determined by the degree of overlap in pollinator assemblages. As long as these assemblages are maintained year after year, hybridization events are likely to be limited. However, one major problem concerning the Portuguese populations of N. cavanillesii is the dramatic change that the surrounding area has experienced because of the construction of the Alqueva Dam and the subsequent artificial water body. This great perturbation will undoubtedly alter the population dynamics of nearby species in the coming years (Ballester-Hernández et al., 2000
). The consequences of this disturbance in pollinator services are difficult to predict but may be substantial (Jennerston, 1988
; Bosch et al., 1998
; Neel, 2002
; Saunders and Sipes, 2006
). Most importantly, this disturbance could provide new suitable habitats for the hybrids (Anderson, 1949
; Heiser, 1979
) and reduce the available habitats of the parental species (Levin et al., 1996
).
Another important issue to consider along with variation of hybridization events in time is the variation of such events in space. In other words, is it possible to generalize these results to other sympatric populations of N. serotinus and N. cavanillesii? The spatial variation of plant interactions (Herrera et al., 2001
) has been neglected in most pollination studies. A recent survey of 664 plant species showed that data had come from multiple populations in only 7.5% of the species and that over 60% of the species had significant geographical variation in pollinators (Herrera et al., 2006
). Therefore, we cannot generalize our results to other populations. Changes in phenological windows, pollinator specificity, and flower constancy may lead to scenarios different from that observed in the Portuguese populations.
Hybridization: extinction of rare species vs. new lines of evolution in the genus
Successful production of viable hybrid seeds in N. cavanillesii, with levels of fruit set and seeds per fruit similar to those found in the intraspecific crosses, could potentially limit its female reproductive fitness. Nevertheless, in the Portuguese populations, we do not expect competition between congeneric and self pollen because these species only share one pollinator and that pollinator has high flower constancy. Similarly, the low incidence of interspecific pollination essentially eliminates the risk of resource competition between developing N. cavanillesii and hybrid seeds in the single fruit produced by N. cavanillesii plants. Although the hybrid's pollen is partially sterile, 40–60% of the hybrids form viable seeds. We do not know the rate of gene flow between the hybrid and the parental species but because of the low abundance of the hybrid, we expect introgression in this population to be low. Therefore, we believe that the hybrid does not currently pose a significant threat to N. cavanillesii in the studied system. In other populations where the factors that limit hybridization may not be so restrictive and the hybrid is more common, the threat of extinction of N. cavanillesii by the hybrid is likely to be greater. In fact, the presence of large hybrid populations in the southeastern Iberian Peninsula in locations where N. cavanillesii is absent may result from displacement of the latter by the hybrid species. Further work is needed in other sympatric locations to determine whether the greater abundance of the hybrid undermines the viability of N. cavanillesii populations.
Various studies illustrate the negative impact of hybrids on the subsistence of rare species (e.g., Levin et al., 1996
; Antilla et al., 1998
; Carney et al., 2000a
; Wolf et al., 2001
). These and other previous experiences have led most conservation biologists to recommend the eradication of the hybrids as well as the prevention of contact between parental populations as principal methods to preserve progenitor populations (Parsons and Hermanutz, 2006
; Smidt et al., 2006
). Nevertheless, hybridization sometimes increases fitness by the adding genetic variability, as in the case of the endemics Pitcarnia albiflos and P. staminea (Wendt et al., 2002
). Even when hybrid zones mainly consist of F1s, rare backcross events may allow the transfer of highly advantageous traits (Cozzolino et al., 2006
). On the other hand, some authors recommend the conservation of hybrids because they represent new evolutionary lineages that are important to preserve (Allendorf et al., 2001
; Allendorf and Luikart, 2007
). In this sense, plant conservationists have traditionally neglected hybrid species even when their natural occurrence is threatened. As far as we know, in the western Mediterranean region, only the last version of the Spanish Red List of Vascular Plants (VV.AA., 2000
) acknowledges the need to protect a hybrid (Phlomis xmargaritae Aparicio & Silvestre), whereas current conservation laws (such as the Endangered Species Act in the USA or the Species and Habitat Directive in Europe) tend to disregard hybrid zones and hybridizing species (Allendorf et al., 2001
). In light of the general negative perception of the role of hybridization in conservation biology, it is important to remember that natural hybridization is a common phenomenon and that hybridization does not always depend on human-mediated habitat disturbance (Cozzolino et al., 2006
). Thus, the high number of natural hybrids described in Narcissus (e.g., Fernandes, 1968
; Blanchard, 1990
) suggests that hybridization may be an important evolutionary process in this genus. In fact, hybridization between N. cavanillesii and N. serotinus has led to the formation of a new viable taxon. Traditionally, the scientific community has been keener to preserve species boundaries than potentially new evolutionary lineages, but this perception may change when more evidence accumulates about the role and importance of hybridization in plant speciation.
When hybrids are formed as a result of natural contact between two congeneric populations, a complex variety of ecological and genetic parameters may influence the risk of extinction of one or both progenitors; these parameters include the vigor and fertility of the hybrids (Levin et al., 1996
; Carney et al., 2000b
), the relative and absolute sizes of the hybridizing populations (Levin et al., 1996
; Ellstrand et al., 1999
), demographic stochasticity, and habitat requirements (Rhymer and Simberloff, 1996
; Wolf et al., 2001
). To date there is no clear predictive framework to assess the likelihood that hybridization will lead to extinction or the speed with which extinction may occur (but see Huxel, 1999
and Wolf et al., 2001
). Given the intrinsic importance of hybrids as new evolutionary pathways, instead of trying to immediately eradicate hybrids whenever a contact affects a rare species, conservation biologists should develop methods to predict likely outcomes of interactions between species. A research program to accomplish this task for Narcissus and other cases of natural hybridization would require (1) specific data on vital rates of the hybrid and parental populations for a minimum of 5 yr, (2) estimation of reproductive fitness of specific and interspecific crossings, (3) a genetic characterization of the hybrid and threatened parental populations so that researchers could infer interspecific gene flux and the degree of genetic introgression, and (4) the development of demographic and genetic models that simulate the trends of the populations under different scenarios (Huxel, 1999
; Wolf et al., 2001
). With further research on the in situ performance of hybrids in many other sympatric congeneric population systems, researchers may gather sufficient data to generalize about the effects of hybridization on other plant species in other environments.
FOOTNOTES
1 The authors thank E. Salvado, S. Albano, and M. J. Albert for field experience in part of this work; L. De Hond for linguistic assistance; and J. Fuertes Aguilar, G. Nieto Feliner, and three anonymous referees, whose comments largely contributed to improving the manuscript. Insect identification was kindly provided by the Natural History Museum of London. This study was promoted by EDIA, S. A., and co-financed by EDIA, S. A., and European Regional Development Funds (ERDF). 
4 Author for correspondence (e-mail: icmarques{at}fc.ul.pt
)
5 Present address: Dpto. Biología Vegetal, Escuela T. S. Ing. Agrónomos, Universidad Politécnica de Madrid, Av. Complutense s/n. 28040 Madrid, Spain
LITERATURE CITED
Albert M. J. Escudero A. Iriondo J. M.. 2001. Female reproductive success of narrow endemic Erodium paularense in contrasting microhabitats. Ecology 82: 1734-1747.[CrossRef][ISI]
Allendorf F. Leary R. F. Spruell P. Wenburg J. K.. 2001. The problems with hybrids: setting conservation guidelines. Trends in Ecology & Evolution 16: 613-622.[CrossRef][ISI]
Allendorf F. Luikart G.. 2006. Conservation and the genetics of populations Blackwell, Malden, Massachusetts, USA.
Anderson E.. 1949. Introgressive hybridization Wiley, New York, New York, USA.
Anderson R. C. Schelfhout S.. 1980. Phenological patterns among tallgrass prairie plants and their implications for pollinator competition. American Midland Naturalist 104: 253-263.[CrossRef][ISI]
Antilla C. K. Daehler C. C. Rank N. E. Strong D. R.. 1998. Greater male fitness of a rare invader (Spartina alterniflora) threatens a common native (Spartina foliosa) with hybridization. American Journal of Botany 85: 1597-1601.
Arnold M. L.. 1992. Natural hybridization as an evolutionary process. Annual Review of Ecology and Systematics 24: 521-524.[CrossRef][ISI]
Arnold M. L.. 1997. Natural hybridization and evolution Oxford University Press, New York, New York, USA.
Ballester-Hernández S. Rosselló-Graell A. Draper D. Correia A. I. D.. 2000. Monitorização de plantas prioritárias na área da albufeira do Alqueva. Linhas metodológicas. Portugaliae Acta Biologica 19: 201-218.
Barrett S. C. H. Harder L. D.. 2005. The evolution of polymorphic sexual systems in daffodils (Narcissus). New Phytologist 165: 45-53.[CrossRef][ISI][Medline]
Barrett S. C. H. Lloyd D. G. Arroyo J.. 1996. Stylar polymorphisms and the evolution of heterostyly in Narcissus. In D. G. Lloyd and S. C. H. Barrett [eds.], Floral biology. Studies on floral evolution in animal-pollinated plants Chapman and Hall, New York, New York, USA.
Barton N. H. Hewitt G. M.. 1985. Analysis of hybrid zones. Annual Review of Ecology and Systematics 16: 113-148.[CrossRef][ISI]
Blanchard J. W.. 1990. Narcissus: a guide to wild daffodils Alpine Garden Society, Surrey, UK.
Bosch M. Simón J. Molero J. Blanché C.. 1998. Reproductive biology, genetic variation and conservation of the rare endemic dysploid Delphinium bolosii (Ranunculaceae). Biological Conservation 86: 57-66.[CrossRef][ISI]
Burke J. M. Carney S. E. Arnold M. L.. 1998. Hybrid fitness in the Louisiana irises: analysis of parental and F1 performance. Evolution 52: 37-43.[CrossRef][ISI]
Campbell D. R. Waser N. M. Pederson G. T.. 2002. Predicting patterns of mating and potential hybridization from pollinator behavior. American Naturalist 159: 438-450.[CrossRef][ISI]
Carney S. E. Gardner K. A. Rieseberg L. H.. 2000a. Evolutionary changes over the fifty-year history of a hybrid population of sunflowers (Helianthus). Evolution 54: 462-474.[CrossRef][ISI][Medline]
Carney S. E. Wolf D. Rieseberg L. H.. 2000b. Hybridization and forest conservation. In T. J. B. Boyle, A. Young, and D. Boshier [eds.], Forest conservation genetics: principles and practice 167-182 CSIRO Publishing, Collingwood, Victoria, Australia.
Charlesworth B. Charlesworth D.. 2000. Reproductive isolation: natural selection at work. Current Biology 10: 68-70.[CrossRef][ISI][Medline]
Cozzolino S. Nardella A. M. Impagliazzo S. Widmer A. Lexer C.. 2006. Hybridization and conservation of Mediterranean orchids: should we protect the orchid hybrids or the orchid hybrid zones?. Biological Conservation 129: 14-23.[CrossRef][ISI]
Ellstrand N. C.. 1992. Gene flow by pollen: implications for plant conservation. Oikos 63: 77-86.[CrossRef][ISI]
Ellstrand N. C. Prentice H. C. Hancock J. F.. 1999. Gene flow and introgression from domesticated plants into their wild relatives. Annual Review of Ecology and Systematics 24: 217-242.[CrossRef]
Ellstrand N. C. Whitkus R. Rieseberg L. H.. 1996. Distribution of spontaneous plant hybrids. Proceedings of the National Academy of Sciences, USA 93: 5090-5093.
Fernandes A.. 1968. Keys to the identification of native and naturalized taxa of the genus Narcissus L. Daffodil Tulip Year Book 59: 37-66.
Font Quer P.. 1927. Notas sobre la flora gaditana. Boletín de la Real Sociedad Española de Historia Natural 27: 39-46.
Graham S. W. Barrett S. C. H.. 2004. Phylogenetic reconstruction of the evolution of stylar polymorphisms in Narcissus (Amaryllidaceae). American Journal of Botany 91: 1007-1021.
Grant V.. 1949. Pollination systems as isolating mechanisms in angiosperms. Evolution 3: 82-97.[CrossRef][ISI][Medline]
Grant V.. 1981. Plant speciation Columbia University Press, New York, New York, USA.
Grant V.. 1994. Modes and origins of mechanical and ethological isolation in angiosperms. Proceedings of the National Academy of Sciences, USA 91: 3-10.
Heiser C. B., Jr.. 1979. Hybrid populations of Helianthus divaricatus and H. microcephalus after 22 years. Taxon 28: 71-75.[CrossRef][ISI]
Herrera C. M. Castellanos M. C. Medrano M.. 2006. Geographical context of floral evolution: towards an improved research programme in floral diversification. In L. D. Harder and S. C. H. Barrett [eds.], Ecology and evolution of flowers 278-294 Oxford University Press, New York, New York, USA.
Herrera C. M. Sanchez-Lafuente A. M. Medrano M. Guitián J. Cerdá X. Rey P. J.. 2001. Geographical variation in autonomous self-pollination levels unrelated to pollinator service in Helleborus foetidus (Ranunculaceae). American Journal of Botany 88: 1025-1032.
Huxel G. R.. 1999. Rapid displacement of native species by invasive species: effects of hybridization. Biological Conservation 89: 143-152.[CrossRef][ISI]
Instituto de Meteorologia.. 1961–1990. Anuários climatológicos de Portugal Instituto de Meteorologia, Lisboa, Portugal.
Jennerston O.. 1988. Pollination in Dianthus deltoides (Caryophyllaceae): effects of habitat fragmentation on visitation and seed set. Conservation Biology 2: 359-366.[CrossRef][ISI]
Johnston J. A. Donovan L. A. Arnold M. L. 2004. Novel phenotypes among early generation hybrids of two Louisiana iris species: flooding experiments. Journal of Ecology 92: 967-976.[CrossRef][ISI]
Levin D. A. Anderson W. W.. 1970. Competition for pollinators between simultaneously flowering species. American Naturalist 104: 455-467.[CrossRef][ISI]
Levin D. A. Francisco-Ortega J. Jansen R. K.. 1996. Hybridization and the extinction of rare plant species. Conservation Biology 10: 10-16.[CrossRef][ISI]
Marques I. Rosselló-Graell A. Draper D.. 2005. Narcissus xperezlarae Font Quer (Amaryllidaceae). A new taxon for the Portuguese flora. Flora Mediterrranea 15: 211-217.
Marques I. Rosselló-Graell A. Draper D. Iriondo J. M.. 2007. Pollination ecology and hybridization between Narcissus cavanillesii A. Barra & G. Lopéz and N. serotinus L. in Portugal. Bocconea: in press.
Mathew B.. 2002. Classification of the genus Narcissus. In G. R. Hanks [ed.], Narcissus and daffodil: the genus Narcissus 30-52 Taylor and Francis, New York, New York, USA.
Mayr E.. 1992. A local flora and the biological species concept. American Journal of Botany 79: 222-238.[CrossRef][ISI]
Neel M. C.. 2002. Conservation implications of the reproductive ecology of Agalinis acuta (Scrophulariaceae). American Journal of Botany 89: 972-980.
Parsons K. Hermanutz L.. 2006. Conservation of rare, endemic braya species (Brassicaceae): breeding system variation, potential hybridization and human disturbance. Biological Conservation 128: 201-214.[CrossRef][ISI]
Pérez Lara J. M.. 1882. Plantarum novarum aliquarum descriptio ad floram gaditanam pertinetium. Anales de la Real Sociedad Española de Historia Natural 11: 399-403.
Pérez-Barrales R. Vargas P. Arroyo J.. 2003. Convergent evolution of flower polymorphism in Narcissus (Amaryllidaceae). New Phytologist 161: 235-252.[ISI]
Pérez-Barrales R. Vargas P. Arroyo J.. 2006. New evidence for the Darwinian hypothesis of heterostyly: breeding systems and pollinators in Narcissus sect. Apodanthi. New Phytologist 171: 553-567.[ISI][Medline]
Rhymer J. M. Simberloff D. S.. 1996. Genetic extinction through hybridization and introgression. Annual Review of Ecology and Systematics 27: 83-109.[CrossRef][ISI]
Rieseberg L. H.. 2006. Hybridization and the conservation of plants. In F. Allendorf and G. Luikart [eds.], Conservation and the genetics of populations, 446 Blackwell, Malden, Massachusetts, USA.
Rosselló-Graell A. Marques I. Draper D.. 2003. Segunda localidad de Narcissus cavanillesii A. Barra & G. López (Amaryllidaceae) para Portugal. Acta Botanica Malacitana 28: 196-197.
Saunders N. E. Sipes S. D.. 2006. Reproductive biology and pollination ecology of the rare Yellowstone Park endemic Abronia ammophila (Nyctaginaceae). Plant Species Biology 21: 75-84.[CrossRef][ISI]
Smidt E. C. Silva-Pereira V. Borba E. L.. 2006. Reproductive biology of two Cattleya (Orchidaceae) species endemic to north-eastern Brazil. Plant Species Biology 21: 85-91.[CrossRef][ISI]
Sokal R. R. Rohlf F. J.. 1995. Biometry, 3rd ed Freeman, New York, New York, USA.
Soler J. X.. 1998. Descubrimiento de Narcissus perezlarae Font Quer (Amaryllidaceae) en el Levante español. Anales del Jardín Botánico de Madrid 56: 165-166.
Stebbins G. L.. 1950. Variation and evolution in plants Columbia University Press, New York, New York, USA.
Templeton A. R.. 1989. The meaning of species and speciation: a genetic perspective. In D. Otte and J. A. Endler [eds.], Speciation and its consequences 3-27 Sinauer, Sunderland, Massachusetts, USA.
Thomson J. D.. 1980. Skewed flowering distributions and pollinator attraction. Ecology 61: 572-579.[CrossRef][ISI]
Torres E. Iriondo J. M. Pérez C.. 2002. Vulnerability and determinants of reproductive success in the narrow endemic Antirrhinum microphyllum (Scrophulariaceae). American Journal of Botany 89: 1171-1179.
VV.AA [various authors].. 2000. Lista roja de flora vascular española (valoración según las categorías UICN). Conservación Vegetal 6: 11-38.
Webb D. A.. 1980. Amaryllidaceae. In T. G. Tutin, V. H. Heywood, N. A. Burges, D. M. Moore, D. H. Valentine, S. M. Walters, D. A. Webb, A. O. Chater, and I. B. K. Richardson [eds.], Flora Europaea, vol. 5 76-77 Cambridge University Press, Cambridge, UK.
Wendt T. Canela M. B. F. Klein D. E. Rios R. I.. 2002. Selfing facilitates reproductive isolation among three sympatric species of Pitcairnia (Bromeliaceae). Plant Systematics and Evolution 232: 201-212.[CrossRef][ISI]
Wolf D. E. Takebayashi N. Rieseberg L. R.. 2001. Predicting the risk of extinction through hybridization. Conservation Biology 15: 1039-1053.[CrossRef][ISI]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
= 0.086, P = 0.092, for Narcissus cavanillesii (N = 301; closed circles) and