Plant Biology ISSN 1435-8603

RESEARCH PAPER

Long-term variation in seed mass and seed production in populations of Paris quadrifolia
ski I. Kosin
sk, Gdansk, Poland Department of Biology and Pharmaceutical Botany, Medical University of Gdan

Keywords Breeding system; long-term seed production; seed mass; seed size number trade-off. Correspondence ski, Department of Biology and I. Kosin Pharmaceutical Botany, Medical University of sk, Al. Gen. J. Hallera 107, 80-416 Gdan Gdansk, Poland. E-mail: gorkos@amg.gda.pl Editor T. Elzenga Received: 26 October 2007; Accepted: 15 January 2009 doi:10.1111/j.1438-8677.2009.00199.x

ABSTRACT Seed production of the perennial herb Paris quadrifolia L. (Liliaceae) was investigated in ve populations in northern Poland. The long-term seed production per square metre differed signicantly among populations and years. Moreover, throughout the 7-year study period, 30% of both whole ripe fruits and seeds alone were predated. Variation in seed mass per fruit in both space and time was signicant. Throughout the 7-year study, nearly all the marked individuals produced fruits once every 2 years. The most frequent break between fruiting was 2 years and the longest was 5 years. Only the mean seed mass in fruits of the same individual varied signicantly over subsequent years. In the ve populations, the number of ovules, number of seeds in the fruit and seed mass varied signicantly between populations. However seed ovule ratio did not differ in fruits in the ve populations. The seed mass number trade-off in fruits was strongly partially correlated when the effect of total seed mass was considered. Breeding experiments suggest that P. quadrifolia has a substantial capacity for both inbreeding and outbreeding. There were no signicant differences in the seed ovule ratio, seed number or seed mass in fruits produced from bagged or control owers. However, both seed ovule ratio and number of seeds were signicantly lower in fruits from emasculated owers.

INTRODUCTION Seed production is a signicant stage in the life history of seed-bearing plants (Harper 1977; Stearns 1992). Seed size is an important trait of the adaptive capability of plants (Stearns 1992; Geritz 1995; Turnbull et al. 1999) and is often also correlated with seedling size and survival (Reader 1993; Moles & Westoby 2004). This correlation can have a signicant impact on recruitment and coloni borg & Eriksson 1997; zation ability of plant species (Fro Turnbull et al. 1999; Jakobsson & Eriksson 2002). The seed size of a given species is less variable than other reproductive traits, such as the number of seeds, owers and fruits or fruit set (Marshall et al. 1986; Silvertown & Charlesworth 2001). Nevertheless, substantial differences in seed size within species have been repeatedly demonstrated, and signicant variations in this trait are apparent among populations (Mendez 1997; Vaughton &
206

ckRamsey 1998; Jacquemyn et al. 2001), individuals (Sto lin & Favre 1994; Eriksson 1999; Sletvold 2002) and individual fruits (Diggle 1995; Mendez 1997; Aniszewski et al. 2001). Both environmental and hereditary factors can inuence seed size (Schmid & Dolt 1994; Jofuku et al. 2005); however, the ability to inherit a given seed size is often low (Silvertown 1989) because the variation within plants and fruits can surpass that among individuals of a given species (Mendez 1997; Jacquemyn et al. 2001). Variation in seed size (offspring) within a species might cklin & Favre result from conicts among offspring (Sto 1994), interactions among parent plants (Ganger 1997; Niesenbaum 1999) or parent offspring conict (Uma Shaanker et al. 1988). Seed size variation can also be explained by selection pressures (Stearns 1992), physiological factors (Diggle 1995), decreased resources during cklin & fruiting and position effects within plants (Sto Favre 1994; Aniszewski et al. 2001). In addition, seed size

Plant Biology 12 (2010) 206214 2009 German Botanical Society and The Royal Botanical Society of the Netherlands

 ski Kosin

Seed mass and seed production of Paris quadrifolia

variability in a given species might be related to population size (Jacquemyn et al. 2001). Location and seasonal and annual effects on seed mass are presumed to be the result of differences in the environment in which the seed developed on the parent plant (Fenner & Thompson 2005). Nevertheless, there is usually no simple dependency between seed size and seed quality on seedling per n (2005) formance. Results obtained by Lehtila & Ehrle suggest that factors that affect seed size in the late stages of seed maturation do not affect seed quality. Paris quadrifolia L., which is common in Europe, is characterized as a self-fertile, mainly autogamous and protogynous plant (Hegi 1909); however, there are few data on the share of outcrossing pollination within the breeding system. Paris quadrifolia is a perennial herb that spreads vegetatively through rhizomes and also reproduces through seeds. The rhizome sends up one shoot per year with a single ower that bears a single multi-seeded fruit with toxic pulp. Paris quadrifolia is used in traditional and homeopathic medicine and contains many biologically active components (Harborne et al. 1999). Breeding system and seed production both affect genotypic diversity in populations, which is important for plants that spread vegetatively (Silvertown & Charlesworth 2001). Moreover, there are few data on the long-

term reproductive dynamics of either perennial plant populations or individuals (e.g. Primack & Stacy 1998; Brzosko 2002; Ohara & Kawano 2005). The aim of the present long-term study of P. quadrifolia was to determine: (i) degree of self- and outcrossing pollination within a breeding system; (ii) quantity of seed production and its dynamics in a population; (iii) long-term variation in seed number and mass in individual plants; and (iv) seed mass number trade-offs in individuals, within populations and among populations. MATERIALS AND METHODS
Study sites and population characteristics

The investigations were conducted in ve populations sk (northern Poland). within a radius of 100 km of Gdan The Swelinia (SW), Grabowiec (GB) and Ryjewo (RW) populations were on the edges of river valleys, while the Bielawy (BL) and Wdzydze (WD) populations were near large lakes. The populations were in forest areas of 0.2 16.0 ha. The characteristics of particular study sites are given in Table 1. Information on soil conditions is based on chemical analyses performed at the Chemical Agriculture Station. The nomenclature of species follows Tutin et al. (19641980).

Table 1. The ve study sites, with information on the area size (ha), average density of Paris quadrifolia (shoots m2), soil conditions and the dominant species of the study sites. average density of P. quadrifolia (shoots m2) 85

site Wdzydze (WD)

area (ha) 0.3

soil conditions organic soil periodically ooded, least acidic and most nutrient rich in phosphorus, potassium, magnesium and nitrogen. fresh, mineralorganic soil, acidic and rich in nitrogen. fresh, mineral, sandy and clay soil, weakly acidic and poor in nitrogen and phosphorus.

dominant species forest stand Alnus glutinosa

ground cover Cardamine amara, Cirsium palustre, Thelypteris palustris

Bielawy (BL)

0.1

189

Quercus petraea, Alnus glutinosa, Corylus avellana Tilia cordata, Carpinus betulus, Quercus robur

Ryjewo (RW)

0.08

21

Grabowiec (GB)

0.04

22

damp, mineral soil of sand and clay, acidic and moderately rich in nitrogen. damp, mineral soil of clay and sand, weakly acidic, moderately rich in magnesium but poor in nitrogen, phosphorus and potassium.

Betula pendula, Carpinus betulus, Alnus glutinosa

Swelinia (SW)

0.005

45

Alnus glutinosa, Fraxinus excelsior, Corylus avellana

Anemone nemorosa, Aegopodium podagraria, Oxalis acetosella Stellaria holostea, Galium odoratum, Lamiastrum galeobdolon, Maianthemum bifolium, Polygonatum multiorum Anemone nemorosa, Stellaria holostea, Oxalis acetosella, Lamiastrum galeobdolon, Hepatica nobilis Anemone nemorosa, Ranunculus caria, Lamiastrum galeobdolon, Aegopodium podagraria

Information on soil conditions was based on chemical analyses performed at the chemical agriculture station.

Plant Biology 12 (2010) 206214 2009 German Botanical Society and The Royal Botanical Society of the Netherlands

207

Seed mass and seed production of Paris quadrifolia

ski Kosin

Seed production

A permanent plot of 2 m2 was established in the BL population to study seed production of individual plants in subsequent years. In 2000 and 2001, 20 and 23 owering shoots, respectively, were marked and the position of each shoot was recorded on a map. As the rhizomes of P. quadrifolia rarely branch, each shoot was considered to be an individual (see Kranczoch 1997). The plot was visited annually over 7 years, and generative or vegetative shoots (i.e. a shoot without fruits) and size (height) of each shoot were recorded. The marked position on each shoot was updated annually by moving it to the next bud of the subsequent years shoot. Each year, the ripe fruits were harvested from each shoot, air-dried, and the seeds removed, counted and weighed individually to the nearest 0.1 mg. Long-term variations in seed production were investigated from 2000 to 2006 in the BL and WD populations. Four permanent plots of 1 m2 were established within each population. In May, when P. quadrifolia was in ower, the number of generative and vegetative shoots was counted in each plot. Every year in July, when the fruits were fully ripe, they were harvested and the height of the fruit-bearing shoots was recorded. Seeds of fruits that had been air-dried were counted and weighed individually. Measurements were performed on 613 shoots (fruits) and 4520 seeds from the permanent plots. Moreover, in each of the ve populations, one representative patch (24 m2) was chosen and ripe fruits were harvested in 2005 and 2006. The seeds and ovules of 316 fruits were recorded. The rate of seed set per fruit (shoot) was calculated as the ratio of the total number of seeds to the total number of ovules. Aborted (soft or unlled) seeds were included in the number of ovules. The airdried seeds were weighed individually.
Breeding system

culated and bagged and 30 were marked and left as a controls; in the BL population, 30 owers were emasculated and 30 were left as controls. In July, ripe fruits from all marked shoots were harvested, and seeds and ovules were counted immediately to evaluate seed set. The seeds were weighed individually after they were air-dried. The rate of fruit set in the population was calculated as the ratio of the number of fruits to the number of owers.
Data analysis

The normality of the frequency distributions of the values was estimated with the ShapiroWilks test. For traits whose distribution was signicantly different from normal (number of ovules, seeds in fruits, seed ovule ratio, seed mass, fruit and seed production per plot), the differences between samples were tested with the MannWhitney (MW) U-test. The signicance of the differences in mean values of shoot size was estimated with the t-test. Differences in shoot height, seed number and seed mass among marked individuals were estimated with the Wilcoxon test for pair-wise comparisons. Seasonal dynamics and variation in fruit and seed production, seed set, number of ovules, seeds and seed mass among populations was analysed with the KruskalWallis (KW) test. The dependencies between shoot height, number of ovules, seed and seed mass were determined with Pearsons correlation and the partial correlation coefcient (Sokal & Rohlf 1995). All statistical analyses were conducted with statistica 7.1 (StatSoft Inc, Tulsa, OK). RESULTS
Seed production and plant development

Experimental studies of breeding systems were conducted in the BL and WD populations in 2005 and 2006. Flowers from randomly chosen generative shoots were marked and treated as follows: (i) owers were bagged in ne mesh nets prior to anthesis to prevent xenogamy; (ii) owers were emasculated prior to pollen dehiscence to prevent autogamy and were left in open-pollination conditions; (iii) owers were emasculated and bagged to test for apomixis; and (iv) owers were left in open-pollination conditions as controls. Thirty-ve owers were bagged, 30 were emasculated and 30 were left as controls in both populations in 2005. In July, the fruits were harvested from shoots, seeds were counted and then weighed individually after they were air-dried. In 2006, the experiment was repeated in these same populations in order to supplement the results with seed set, and also because of the high losses of marked shoots in population WD in 2005 due to roe deer foraging. In the WD population, 30 owers were bagged, 30 were emasculated, 12 were emas208

Fruit-bearing individuals were chosen and marked at the beginning of the study. In each year, the rhizomes produced aboveground shoots that were either owering (and fruit-bearing) or vegetative. A 1-year break in fruit production occurred in 23% of individuals, but a 2-year break was more frequent (42%). The longest (and rarest) fruit production break was 5 years (10%). Throughout the 7-year study, nearly all individuals produced fruits once every 2 years. However, in subsequent years fruiting at periods of 35 years occurred in 12% of individuals. Fruit of the same individual in two subsequent years differed signicantly only with regard to mean seed mass (Table 2). The number of seeds, total seed mass and height of maternal shoots did not differ signicantly. Individuals that produced fruit with a break of 1 year or longer only differed signicantly in seed mass, similar to those that produced fruits in subsequent years (Table 2). Shoots of individuals in the vegetative stage were signicantly shorter than those that had produced fruit in the previous season (Wilcoxon test, P < 0.001). The return of individuals to the generative phase in subsequent year was referred to as the return to the previous size. Breaks of a year in the occurrence of aboveground shoots of

Plant Biology 12 (2010) 206214 2009 German Botanical Society and The Royal Botanical Society of the Netherlands

 ski Kosin

Seed mass and seed production of Paris quadrifolia

Table 2. Characteristics of fruit-producing shoots of marked individuals in the Bielawy (BL) population compared in two subsequent years (top) and after a break of 1 year or longer (bottom). trait shoot height (cm) seed number mean seed mass (mg) total seed mass (mg) shoot height (cm) seed number mean seed mass (mg) total seed mass (mg) n 21 21 21 21 11 11 11 11 rst year 25.5 14.2 4.63 65.7 22.7 15.8 4.32 66.7 2.0 4.4 0.73 22.2 2.9 6.6 0.71 26.5 next year 25.9 16.6 4.02 65.8 2.5 5.1 0.55 22.0 23.0 13.3 3.79 49.5 2.1 5.1 0.67 18.4 after 2 years T; P 87.5; 0.51 61; 0.06 35; 0.005 115; 0.99 31; 0.86 23.5; 0.39 10; 0.041 18; 0.18

Mean and standard deviation; Wilcoxon test for pair-wise comparisons.

marked individuals were also noted. In these instances, vegetative shoots were the rst to develop following the break. Fruit predation occurred each year with a mean of 25%. In 2006, 11 individuals were lost.
Long-term seed production

There was substantial variation in the number of fruitproducing shoots in the permanent plots during the 7-year study. The number of fruits obtained from 1 m2 in the BL and WD populations ranged from 5 to 120 and 0 to 64, respectively. The long-term seed production per square metre differed signicantly between the BL and WD populations, with an average of 223 and 143 seeds from 418 and 195 fruits, respectively (MW test, P < 0.05). Seasonal variation in seed production was signicant in the BL and WD populations (KW test, P < 0.01). However, differences in seed production for both populations were only signicant in 2001, when compared to 2004 and 2005 (Fig. 1). The long-term dynamics of seed production resulted mainly from uctu-

ations in the number of fruit-bearing shoots in the populations (KW test, P < 0.01). However, throughout the study, whole ripe fruits and seeds were predated by small rodents. In both populations, predation affected 30% of fruits, but this varied substantially between years (e.g. all fruits in population WD were predated in 2004). In the 20002006 period, there were 15.0 4.5 (SD) seeds per fruit in the BL population, while in the WD population there were signicantly more seeds per fruit (20.6 6.8; MW test, P < 0.001). The variation in number of seeds per fruit in subsequent years was signicant in both populations (KW test, P < 0.001). In years when the smallest number of fruit-bearing shoots was noted in both populations, fruits had the lowest number of seeds (Fig. 2). The frequency distribution of the number of seeds in fruit was normal in the BL population (skewness = 0.06), while in the WD population it was weakly skewed (skewness = 0.68). The number of seeds per fruit was correlated with shoot height in both the BL and WD populations (r = 0.5 and 0.61, respectively; P < 0.001).

Fig. 1. Seasonal changes in annual seed production (mean SD per square metre; n = 4) in the Bielawy (BL; open bars) and Wdzydze (WD; shaded bars) populations of P. quadrifolia. Bars with the same letter do not differ over all years and the two populations (Kruskal Wallis test, P > 0.05).

Fig. 2. Seed number per fruit (open circles) and seed mass (closed circles) in fruit of P. quadrifolia in the Bielawy (BL) population (n = 418). Data points (mean SD) within each trait with the same letter do not differ (KruskalWallis test, P > 0.05). Please note that the y-axis is not continuous.

Plant Biology 12 (2010) 206214 2009 German Botanical Society and The Royal Botanical Society of the Netherlands

209

Seed mass and seed production of Paris quadrifolia

ski Kosin

Seeds in the BL population were heavier than those in WD (3.77 0.75 and 3.29 0.68 mg, respectively; MW test, P < 0.001). Seed mass distribution in the BL population was slightly skewed ()0.24) and similar to the WD population ()0.32). Annual variation in seed mass in fruit was also signicant in subsequent years (KW test, P < 0.001). Mean seed mass per fruit was weakly correlated with shoot height in both BL and WD populations (r = 0.2 and 0.23, respectively; P < 0.05). Seed mass number trade-off in fruit in both the BL population (r = )0.11; P < 0.05) and together with the WD population (r = )0.25; P < 0.001) was weak. The partial correlation, when total seed mass in the fruit was considered, was always strong (r = )0.9; P < 0.001).
Seed number and seed mass between populations

In the ve populations, the number of ovules per ower ranged from 10 to 77 (mean 30.1 9.0). The number of ovules differed between populations (KW test, P < 0.001) and was weakly correlated with shoot height (r = 0.32; P < 0.001). The number of seeds in the fruit ranged from 2 to 66 (mean 19.9 7.7). Moreover, fruits with a lower seed number when compared to the arithmetic mean were frequently found in all ve populations (Fig. 3). Seed mass ranged from 0.8 to 7.9 mg (mean 3.92 1.05 mg), and lighter seeds (when compared to average mass) were also frequent in all ve populations (Fig. 4). Seed set ranged from 36 to 97% (mean 78.5%). The number and mean mass of seeds in fruits differed between populations (KW test, P < 0.001; Fig. 5); however, seed set did not differ in the ve populations (KW test, P = 0.08; Fig. 5). Shoots from the RW, GB and SW populations were signicantly taller than those from the WD population (t-test, P < 0.001). In all ve populations, shoot height

Fig. 4. Frequency distribution of seed mass of P. quadrifolia. The distribution was based on 6138 individually weighed seeds from ve populations. Seed mass was 3.92 1.05 mg (mean SD) and skewness was 0.42.

Fig. 5. Seed set (asterisks), seed number per fruit (open circles) and seed mass (closed circles) in fruit of P. quadrifolia from ve populations: WD Wdzydze (n = 74), BL Bielawy (n = 43), RW Ryjewo (n = 46), GB Grabowiec (n = 80) and SW Swelinia (n = 73). Data points (mean SD) within each trait with the same letter do not differ (KruskalWallis test, P > 0.05). Please note that the y-axis is not continuous.

Fig. 3. Frequency distribution of seed number in fruit of P. quadrifolia. The distribution was based on 309 fruits from ve populations. Seed number per fruit was 19.9 7.7 (mean SD) and skewness was 1.32.

was positively correlated with mean seed mass in the fruit (r = 0.52; P < 0.001) but less so with the number of seeds in the fruit (r = 0.26; P < 0.001), while seed set was not signicantly correlated with shoot size (P = 0.7). Seed mass number trade-off was only signicant in fruit from the RW population (r = )0.34; P < 0.05). However, when the effect of total seed mass per fruit was considered, the partial correlation was high in all ve populations (r = )0.9; P < 0.001). Moreover, seed mass number trade-off was at an inter-population level, namely fruit with a greater mean number of seeds had

210

Plant Biology 12 (2010) 206214 2009 German Botanical Society and The Royal Botanical Society of the Netherlands

 ski Kosin

Seed mass and seed production of Paris quadrifolia

Fig. 6. Seed set (mean SD) of P. quadrifolia in the Wdzydze (WD) population in untreated control plants (n = 36) and after various treatments: bagged (n = 27), emasculated (n = 26), emasculated and bagged (n = 12). Bars with the same letter do not differ (Kruskal Wallis test, P > 0.05).

pollination in production of seeds in this plant (Fig. 6). Fruit produced from emasculated owers contained 8.8 7.1 seeds weighing 3.32 0.98 mg in the BL population and 13.2 10.5 seeds weighing 4.14 0.64 mg in the WD population, and the seed ovule ratios were 28.2% and 42.3%, respectively. Both seed set and the number of seeds were signicantly lower in comparison with the control (MW test, 0.0001 < P < 0.02). Emasculated and bagged owers did not produce fruits; therefore, it is unlikely that P. quadrifolia is apomictic (Fig. 6). This suggests cross-pollination either by wind or insects in seed production. Paris quadrifolia appears to have a substantial capacity for both inbreeding and outbreeding, and both play important roles in seed production. In addition, fruit set for bagged and control owers was 100%, unlike that in emasculated owers (21.7% in BL and 83.9% in WD populations). DISCUSSION
Breeding system

lighter seeds and vice versa, e.g. in populations SW and RW, respectively (Fig. 5).
Breeding system

In 2005, fruit from control shoots (open-pollination) contained 15.9 6.2 seeds, weighing 4.07 0.59 mg in the BL population and 20.7 4.0 seeds weighing 3.73 0.51 mg in the WD population. Fruit from bagged owers contained 16.4 3.9 seeds, weighing 3.94 0.8 mg and 22 6.9 seeds weighing 4.18 0.59 mg in BL and WD populations, respectively. Differences in the number and mean mass of seeds between treatment and control groups were insignicant (MW test, 0.56 < P < 0.91). Fruit produced by emasculated owers contained 7.9 5.9 seeds, weighing 4.25 0.7 mg in the BL population and 13.8 8.3 seeds weighing 3.57 0.57 mg in the WD population. The number of seeds in fruit of emasculated owers was signicantly lower than in control shoots (MW test, P < 0.01). These differences were not signicant in the WD population, probably due to animal damage to emasculated shoots in this population. The experiment was repeated in both populations in 2006 and data on the number of ovules were included. Control shoots in the BL population had 28.5 8.9 ovules and fruit had 19.2 7.3 seeds weighing 3.66 0.46 mg, while a shoot in the WD population had 31.4 9.6 ovules and 24.6 8.4 seeds weighing 3.79 0.73 mg; the seed ovule ratios were 67.6% and 78.8%, respectively. Fruit produced from bagged owers in the WD population had 22.4 8.5 seeds weighing 3.85 0.52 mg and a seed ovule ratio of 75.5%. These values were similar to those of the control (MW test, 0.5 < P < 0.9), which conrms the dominant role of self-

The owers of P. quadrifolia are self-fertile and protogynous, with a long period in which stigmas are able to accept pollen; they are also visited by insects (Hegi 1909). The bagging and emasculating experiments indicate that self-pollination in the production of seeds is substantially higher than cross-pollination. Cross-pollination here refers only to pollination with pollen from another ower, not necessarily from a different genetic individual. During the investigation, insects were not frequently observed visiting P. quadrifolia owers, which leads to the assumption that wind plays an important role in the transmission of pollen, especially over small distances between owering shoots within patches. Seed set from emasculated owers was highly variable, with values similar to those in the control group (8090%). Nevertheless, a signicant number of emasculated owers did not set fruit. Species in the genus Trillium, which is related to Paris, exhibit variations in breeding systems. Trillium apetalon (also with inconspicuous owers) has a similar breeding system to that of P. quadrifolia; where no seeds are produced in emasculated owers because of a lack of non-self pollen, but articial cross-pollination leads to production of seeds (Ohara et al. 2001). Trillium erectum and T. grandiorum are usually cross-pollinated although they can also set fruit after self-pollination and are not apomictic (Irwin 2000). Trillium reliquum can mature fruits equally well with either cross- or self-pollination and by apomixis (Heckel & Leege 2007). Differences in breeding systems can occur among populations of a given species, such as in T. camschatchense (Ohara et al. 1996), where plants in one region were self-incompatible and had high genetic variability, while in another region they were self-compatible with low genetic variability. The mixed breeding system in the investigated populations of P. quadrifolia probably promotes high genetic diversity.
211

Plant Biology 12 (2010) 206214 2009 German Botanical Society and The Royal Botanical Society of the Netherlands

Seed mass and seed production of Paris quadrifolia

ski Kosin

Clonal diversity within the P. quadrifolia populations investigated by Jacquemyn et al. (2006) was relatively high in comparison with other clonal species.
Variation in seed number and seed mass in populations

Fruit set in the investigated populations of P. quadrifolia with mixed breeding systems was mostly 100%, unlike the obligatory outcrosser populations of T. camschatcense (Tomimatsu & Ohara 2002). Nevertheless, seed production per unit area in the investigated P. quadrifolia populations varied signicantly over a long period. Although stochastic changes in the number of fruits (seeds) were caused by random events (including pre-dispersal seed predation), they were synchronous in both BL and WD n & Eriksson (1993) reported that in populations. Ehrle some patches of P. quadrifolia up to 80% of fruits were removed by small rodents. The signicant decrease in seed number in fruit observed in one season was probably caused by predation of larger fruits with numerous seeds. Moreover, variations in numbers of fruit-bearing shoots were caused by animals (i.e. voles, roe deer), which browse on buds or young shoots, as well as (probably) dormancy of buds. Analysis of data from the 7-year study of seed numbers in fruit from two populations of P. quadrifolia indicates that population variation was higher than variation between years. It is likely that the mixed type breeding system leads to low seasonal variation in the number of seeds in a fruit; whereas, outcrossing populations of T. camschatcense, which differ markedly in seed number per fruit, did not differ between years (Tomimatsu & Ohara 2002). Signicant differences in number of seeds per fruit have been reported for P. quadrifolia populations in various environments (Jacquemyn et al. 2005). Similarly, in wetter environments, fruit contained more seeds (population WD, SW and GB), while there were less seeds fruit in drier environments with a higher population density (BL population; Table 1, Fig. 5). Moreover, these differences were observed throughout the 7-year study period. Mean seed mass in fruit was positively correlated with shoot height in all populations. Fruits from RW, GB and SW had heavier seeds than the WD population (Fig. 5). Shoots from populations on mineral soil that was poor in nitrogen, phosphorus and potassium (RW, GB and SW) grew at a low density and their shoots were signicantly taller than those from the WD population at a higher density on organic soil rich in these elements. Therefore, specic environmental conditions, such as density of shoots, could explain seed mass variations between populations. Parental nutrient content is known to affect seed size (Fenner & Thompson 2005). Populations in habitats richer in nutrients contained taller plants that produced fruits with heavier seeds. The heaviest seeds of P. quadrifolia were noted in the RW population, which had the tallest shoots of all the populations (but this habitat was not nutrient-rich). A similar dependence has been noted
212

gren 1989; Sto cklin & Favre 1994; in other species (A Sletvold 2002). Seed size is the least variable species trait (Marshall et al. 1986). Differences in survival of seeds of varied sizes means that there is an optimal seed mass (Smith & Fretwell 1974). The mass of the smallest seed of P. quadrifolia was 10-fold smaller than the heaviest seed (Fig. 4). Similar, substantial seed mass variation has been reported for many species (e.g. Mendez 1997; Eriksson 1999; Jacquemyn et al. 2001). The trade-off theory of resource allocation suggests that reproduction imposes a measurable cost on the plant. This cost can manifest as a decrease in vegetative growth of parent plants or lower future reproductive ability (Primack & Stacy 1998; Fenner & Thompson 2005). Among the marked P. quadrifolia shoots, the cost of setting fruit in a given year or every 2 years manifested as lower shoot height and or lack of fruit in the next season. However, data on natural variation in fruit production in the rhizomatous perennial Lathyrus vernus provided no evidence that reproduction occurred at any cost to the parent plant n & van Groenendael 2001). (Ehrle
Seed size and number trade-off

The seed size number trade-off is a result of allocation of a relatively xed proportion of plant resources to seeds (Shipley & Dion 1992), and occurs at an intraspe gren 1989; Vaughton & cic level in many species (A Ramsey 1998; Aniszewski et al. 2001). This dependence was observed within both individuals and populations. In single marked individuals of P. quadrifolia, an increase in the number of seeds per fruit led to a decrease in mean seed mass in the subsequent year, while total seed mass in fruits remained similar. This dependence is probably reected in the seed size number trade-off at the interpopulation level: in populations with a higher mean number of seeds per fruit, the mean seed mass was lower. Pearsons correlation between seed size and number within a fruit was often statistically insignicant, except when total seed mass in the fruit was controlled for partial correlation (see Eriksson 1999). Seed size and seed number is not the only factor that shapes seedling establishment and survival (Lehtila & n 2005; Moles & Westoby 2006). Long-term studies Ehrle suggest that the distribution of long-lived forest herbs is n et al. 2006). Paris limited by seed availability (Ehrle quadrifolia has limited dispersal ability (Brunet & von Oheimb 1998; Honnay et al. 1999). Moreover, the cause of weak recruitment of juveniles of this plant was not determined (Baskin & Baskin 2001). Knowledge of seed size and number variation in different populations provided by the current study will be applied to investigate to what extent differences in seed size and environmental conditions inuence seedling establishment and survival of juvenile stages of P. quadrifolia in different habitats.

Plant Biology 12 (2010) 206214 2009 German Botanical Society and The Royal Botanical Society of the Netherlands

 ski Kosin

Seed mass and seed production of Paris quadrifolia

ACKNOWLEDGEMENTS This research was supported by the Medical University of sk. I would also like to express my deep gratitude to Gdan ska for her help with eldwork. Nina Kosin REFERENCES
gren J. (1989) Seed size and number in Rubus chamaemoA rus between-habitat variation, and effects of defoliation and supplemental pollination. Journal of Ecology, 77, 10801092. Aniszewski T., Kupami M.H., Leinonen J. (2001) Seed number, seed size and seed diversity in Washington lupin (Lupinus polyphyllus Lindl.). Annals of Botany, 87, 7782. Baskin C.C., Baskin J.M. (2001) Seeds. Ecology, Biogeography and Evolution of Dormancy and Germination. Academic Press, San Diego. Brunet J., von Oheimb G. (1998) Migration of vascular plants to secondary woodlands in southern Sweden. Journal of Ecology, 86, 429438. Brzosko E. (2002) Dynamics of island populations of Cypripedium calceolus in the Biebrza River valley (northeast Poland). Botanical Journal of the Linnean Society, 139, 6777. Diggle P. (1995) Architectural effects and the interpretation of patterns of fruit and seed development. Annual Review of Ecology and Systematics, 26, 531552. n J., Eriksson O. (1993) Toxicity in eshy fruits a nonEhrle adaptive trait? Oikos, 66, 107113. n J., van Groenendael J. (2001) Storage and the delayed Ehrle costs of reproduction in the understorey perennial Lathyrus vernus. Journal of Ecology, 89, 237246. n J., Mu nzbergova Z., Diekmann M., Eriksson O. (2006) Ehrle Long-term assessment of seed limitation in plants: results from an 11-year experiment. Journal of Ecology, 94, 1224 1232. Eriksson O. (1999) Seed size variation and its effect on germination and seedling performance in the clonal herb Convallaria majalis. Acta Oecologica, 20, 6166. Fenner M., Thompson K. (2005) The Ecology of Seeds. Cambridge University Press, Cambridge. borg H., Eriksson O. (1997) Local colonization and extincFro tion of eld layer plants in a deciduous forest and their dependence upon life history features. Journal of Vegetation Science, 8, 395400. Ganger M.T. (1997) The inuence of pollen addition and ramet isolation on current sexual reproduction in a clonal herb. Oecologia, 110, 231236. Geritz S.A.H. (1995) Evolutionary stable seed polymorphism and small-scale spatial variation in seedling density. American Naturalist, 146, 685707. Harborne J.B., Baxter H., Moss G.P. (1999) Phytochemical Dictionary: A Handbook of Bioactive Compounds from Plants. Taylor and Francis, London. Harper J.L. (1977) Population Biology of Plants. Academic Press, London.

Heckel C.D., Leege L.M. (2007) Life history and reproductive biology of the endangered Trillium reliquum. Plant Ecology, 189, 4957. Hegi G.. (1909) Illustrierte Flora von Mitteleuropa, Vol. 2 1. nchen. Lehman Verlag, Mu Honnay O., Hermy M., Coppin P. (1999) Impact of habitat quality on forest plant species colonization. Forest Ecology and Management, 115, 157170. Irwin R.E. (2000) Morphological variation and female reproductive success in two sympatric Trillium species: evidence for phenotypic selection in Trillium erectum and Trillium grandiorum (Liliaceae). American Journal of Botany, 87, 205214. Jacquemyn H., Brys R., Hermy M. (2001) Within and between plant variation in seed number, seed mass and germinability of Primula elatior: effect of population size. Plant Biology, 3, 561568. n-Ruiz I. Jacquemyn H., Brys R., Honnay O., Hermy M., Rolda (2005) Local forest environment largely affects belowground growth, clonal diversity and ne-scale spatial genetic structure in the temperate deciduous forest herb Paris quadrifolia. Molecular Ecology, 14, 44794488. n-Ruiz I. Jacquemyn H., Brys R., Honnay O., Hermy M., Rolda (2006) Sexual reproduction, clonal diversity and genetic differentiation in patchily distributed populations of the temperate forest herb Paris quadrifolia (Trilliaceae). Oecologia, 147, 434444. Jakobsson A., Eriksson O. (2002) Seed size and frequency patterns of understory plants in Swedish deciduous forests. Ecoscience, 9, 7478. Jofuku K.D., Omidyar P.K., Gee Z., Okamuro J.K. (2005) Control of seed mass and seed yield by the oral homeotic gene APETALA2. Proceedings of the National Academy of Sciences USA, 102, 31173122. nKranczoch J. (1997) Struktur und Dynamik in dichten Besta den von Paris quadrifolia (Trilliaceae). J. Cramer, Berlin. n J. (2005) Seed size as an indicator of seed Lehtila K., Ehrle quality: a case study of Primula veris. Acta Oecologica, 28, 207212. Marshall D.L., Levin D.A., Fowler N.L. (1986) Plasticity of yield components in response to stress in Sesbania macrocarpa and Sesbania vesicaria (Leguminosae). American Naturalist, 127, 508521. Mendez M. (1997) Sources of variation in seed mass in Arum italicum. International Journal of Plant Sciences, 158, 298305. Moles A.T., Westoby M. (2004) Seedling survival and seed size: a synthesis of the literature. Journal of Ecology, 92, 372383. Moles A.T., Westoby M. (2006) Seed size and plant strategy across the whole life cycle. Oikos, 113, 91105. Niesenbaum R.A. (1999) The effects of pollen load size and donor diversity on pollen performance, selective abortion, and progeny vigor in Mirabilis jalapa (Nyctaginaceae). American Journal of Botany, 86, 261269. Ohara M., Kawano S. (2005) Life-history monographs of Japanese plants. 2: Trillium kamtschaticum Ker-Gawl (Trilliaceae). Plant Species Biology, 20, 7582.

Plant Biology 12 (2010) 206214 2009 German Botanical Society and The Royal Botanical Society of the Netherlands

213

Seed mass and seed production of Paris quadrifolia

ski Kosin

Ohara M., Takeda H., Ohno Y., Shimamoto Y. (1996) Variations in the breeding system and the population genetic structure of Trillium camschatchense (Liliaceae). Heredity, 76, 476484. Ohara M., Takada T., Kawano S. (2001) Demography and reproductive strategies of a polycarpic perennial, Trillium apetalum (Trilliaceae). Plant Species Biology, 16, 209217. Primack R., Stacy E. (1998) Cost of reproduction in the pink ladys slipper orchid (Cypripedium acaule, Orchidaceae): an eleven-year experimental study of three populations. American Journal of Botany, 85, 16721679. Reader R.J. (1993) Control of seedling emergence by ground cover and seed predation in relation to seed size for some old-eld species. Journal of Ecology, 81, 169175. Schmid B., Dolt C. (1994) Effects of maternal and paternal environment and genotype on offspring genotype in Solidago altissima L. Evolution, 48, 15251549. Shipley B., Dion J. (1992) The allometry of seed production in herbaceous angiosperms. American Naturalist, 139, 467483. Silvertown J. (1989) The paradox of seed size and adaptation. Trends in Ecology and Evolution, 4, 2426. Silvertown J.W., Charlesworth D. (2001) Introduction to Plant Population Biology. Blackwell, Oxford. Sletvold N. (2002) Effects of plant size on reproductive output and offspring performance in the facultative biennial Digitalis purpurea. Journal of Ecology, 90, 958966.

Smith C.C., Fretwell S.D. (1974) The optimal balance between size and number of offspring. American Naturalist, 108, 499506. Sokal R.R., Rohlf F.J. (1995) Biometry. Freeman, New York. StatSoft, Inc. (2005) statistica, ver. 7.1. http://www.statsoft.com. Stearns S.C. (1992) The Evolution of Life Histories. Oxford University Press, New York. cklin J., Favre P. (1994) Effects of plant size, morphologSto ical constraints on variation in reproductive components in two related species of Epilobium. Journal of Ecology, 82, 735746. Tomimatsu H., Ohara M. (2002) Effects of forest fragmentation on seed production of the understory herb Trillium camschatcense. Conservation Biology, 16, 12771285. Turnbull L.A., Rees M., Crawley M.J. (1999) Seed mass, the competition colonization trade-off: a sowing experiment. Journal of Ecology, 87, 899912. Tutin T.G., Heywood V.H., Burges N.A., Moore D.M., Valentine D.H., Walters S.M., Webb D.A., (eds.) (19641980) Flora Europaea. Cambridge University Press, Cambridge. Uma Shaanker R., Ganeshaiah K.N., Kamaljit S.B. (1988) Parent-offspring conict, sibling rivalry, brood size patterns in plants. Annual Review of Ecology and Systematics, 19, 177205. Vaughton G., Ramsey M. (1998) Sources, consequences of seed mass variation in Banksia marginata (Proteaceae). Journal of Ecology, 86, 563573.

214

Plant Biology 12 (2010) 206214 2009 German Botanical Society and The Royal Botanical Society of the Netherlands

Copyright of Plant Biology is the property of Blackwell Publishing Limited and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.