| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
What's this? |
Ecology |
Department of Botany, Stockholm University, SE-106 91 Stockholm, Sweden
Received for publication December 1, 2006. Accepted for publication June 25, 2007.
ABSTRACT
Timing of seasonal plant development can affect biotic interactions and plant fitness. Phenology is governed largely by temperature and may therefore be affected by global climate warming, making this an important area of research. Several factors in addition to temperature may cause differences in phenology. We studied the influence of local environment, plant size, and reproductive effort on shoot emergence and flowering time of 290 individuals of Actaea spicata (Ranunculaceae), distributed among 25 plots in four populations. We used multiple regression and structural equation models (SEM) to study causal relationships. Among plots, soil temperature and canopy cover explained 63% of the variation in shoot emergence. Soil temperature, slope, and canopy cover together explained 83% of the variation in flowering time. Within plots, small plants on steep south-facing slopes with high soil potassium concentrations emerged earlier in the year. Plants emerging earlier flowered earlier, but no environmental factors affected flowering time directly. We found no effects of reproductive effort. Our results support the view that flowering time of temperate forest herbs is constrained by several environmental factors acting indirectly through effects on shoot emergence time.
Key Words: Actaea spicata • flowering time • forest herb • microenvironment • potassium • shoot emergence • structural equation model • Ranunculaceae • within-population
Plant phenology, and in particular flowering time, has been the focus of many studies ever since the emergence of botany as a science (Schnelle, 1955
; Lieth, 1974
; Rathcke and Lacey, 1985
; and references therein). Recently, causes and consequences of the timing of seasonal plant development have received increased attention in the context of global climate warming. There is evidence of flowering time being strongly influenced by climate, and this may have large ecosystem consequences due to other organisms' dependency on plants (Fitter and Fitter, 2002
). Detailed plant phenology models may be needed to predict consequences of climate change (cf. Beaubien and Hall-Beyer, 2003
). However, a range of factors may be influential, and the ecological and evolutionary causes of variation in flowering time remain to be fully understood (Marquis, 1988
; Ollerton and Lack, 1992
; Debussche et al., 2004
; Goulart et al., 2005
).
Flowering time can affect plant fitness (e.g., Schemske et al., 1978
; Marquis, 1988
; Mahoro, 2002
). Temporal variation in availability of pollinators, seed dispersers, and herbivores may be strong selection pressures on flowering phenology (Augspurger, 1981
; Rathcke and Lacey, 1985
; Evans et al., 1989
; Brody, 1997
). Environmental factors influencing flowering time may thus contribute to spatial variation in plant–pollinator, plant–seed disperser, and plant–seed predator interactions. There are also environmental constraints, e.g., temperate forest herbs that flower in early spring must take advantage of the brief period suitable for photosynthesis between thawing and closing of the tree canopy (Schemske et al., 1978
). In addition, flowering time may be affected by factors influencing vegetative development and time needed for seed and fruit maturation (Diggle, 1999
; Sola and Ehrlén, 2007
).
Ecological causes of variation in flowering time include differences in temperature, photoperiod, and moisture (Rathcke and Lacey, 1985
). In temperate regions, temperature sum models, based on cumulative temperature sums above a threshold value, have proven reliable in predicting flowering time (Diekmann, 1996
; but see Wang, 1960
; and Schemske et al., 1978
). Flowering time also differs between plants growing in different soils. This may result from differences in moisture that cause differences in temperature (Schnelle, 1955
). However, differences in flowering time may also be due to differences in nutrient concentrations (Wielgolaski, 2001
). Flowering phenology may also vary between individuals over very small spatial scales (Marquis, 1988
; Goulart et al., 2005
). Relatively little is known of the role of local plant environment, but part of within-population variation is caused by microclimatic differences (Jackson, 1966
; Inouye et al., 2003
). However, possible effects of other environmental factors, such as soil nutrients, have not received much attention. Furthermore, variation in phenology can be caused by genetic differences as well as the resource state of the plant (McMillan and Pagel, 1958
; Wang, 1960
). Resource state, in turn, may be correlated with age, size, and reproductive effort in the previous years (Schemske et al., 1978
).
In this study we investigate to what extent differences in local environment, plant size, and previous reproductive effort is related to within-population variation in vegetative and flowering phenology of Actaea spicata. The environment is described by canopy cover, plant density, the slope of the ground, and the following soil characteristics: depth, temperature, water content, pH, nutrient concentrations, and litter depth. Actaea spicata is a forest herb that flowers in early summer and that occurs in temperate regions with large seasonal changes in environmental conditions. We seek to answer the following questions: (1) Is within-population variation in flowering time of A. spicata related to differences in local environment, plant size, and reproductive effort in the previous year? (2) Do environmental factors act directly on flowering time or indirectly through vegetative phenology? (3) Do patterns change with spatial scale?
MATERIALS AND METHODS
Study area and species
The study was conducted on four patches (hereafter referred to as populations) of A. spicata in the forests of Tullgarn natural reserve, 45 km SSW of Stockholm, Sweden (58°6' N, 17°4' E). Over the four study sites, the tree flora ranged from pure deciduous stands, consisting mainly of Quercus robur and Corylus avellana, to coniferous stands dominated by Picea abies. Canopy cover ranged from relatively open to almost completely closed. Actaea spicata was abundant at all study sites. We are also currently studying the effect of the environment on the demography of A. spicata in this area (J. P. Dahlgren and J. Ehrlén, unpublished data).
Actaea spicata L. is distributed over Scandinavia and central and eastern Europe (Hultén and Fries, 1986
). It is a long-lived herb occurring in shady, well-drained habitats in rich deciduous and coniferous forests, often on limestone (Pellmyr, 1984
). In the study area, aerial parts become visible around the first of May. Vegetative growth stops and flowering begins approximately 3 wk after shoot emergence. The growth form is typical of an early-summer flowering woodland herb, i.e., A. spicata is taller than spring ephemerals and has an umbrella-like leaf display (cf. Givnish, 1987
). Inflorescences have 1–30 white flowers (J. P. Dahlgren, personal observation). Flowering may extend into July, and individual shoots sometimes produce up to four inflorescences (Eriksson, 1995
). Potential pollinators include several insect orders, but most reported visitors belong to Coleoptera and Diptera (Pellmyr, 1984
). Actaea spicata is either self-incompatible, but facultatively apomictic, or able to self-pollinate; in either case, seed set is nearly 100% (Pellmyr, 1984
; Eriksson, 1995
). The rhizome may split, but reproduction is predominantly sexual (J. P. Dahlgren, personal observation). The fruit is a black berry, ripening in August. Within flowering shoots, seeds of early infructescences are often eaten by the larvae of Eupithecia immundata (Geometridae), but late infructescences completely escape seed predation (Eriksson, 1995
; von Zeipel et al., 2006
). The main seeds dispersers are thought to be rodents (Eriksson, 1994
).
Data collection
We observed 290 individuals flowering in 2005. Plants were distributed among five to seven 5 x 5 m plots (25 plots in total), in four populations. Distances between populations ranged from 0.8 to 2 km. Distances between plots in a population were 0–50 m. The location of each individual was mapped, and the number of fruits was recorded for all plants in 2004. In 2005, dates of shoot emergence and start of flowering were determined. The height of each plant was recorded once between 30 April and 7 May, when most individuals had visible aerial parts. Each population was visited two or three times, and all height measures were recalculated to represent the height on 6 May by using linear inter- or extrapolation within a population. Shoot elongation is rapid in A. spicata, and the maximum shoot height is reached within weeks of emergence. The ratio of height on 6 May to the height of the fully developed shoot, measured in July, was therefore used as an estimate of vegetative phenology, hereafter referred to as shoot emergence. The mean height ratio was 0.22 and the variance 0.025. The flowering stage of the first inflorescences on all individuals was recorded twice during bloom. Number of flowers in each of the following categories was counted: undeveloped buds, fully developed buds, fresh open flowers, flowers in which the majority of stamens had shed their pollen, and flowers in which most stamens had fallen off. The relative time of a flower in each category was calculated. For each inflorescence, interpolation between the two observations was used to determine the date when 50% of the flowers of an inflorescence had been in the fresh open category, hereafter called flowering time. The mean Julian day of flowering was 151.4, and the variance was 13.74. Plant height and the diameter at the base of each shoot were measured in July. Plant size was calculated by summing the products of shoot diameter to the power of two multiplied by height. We used the summed sizes of conspecifics within a circle of 1 m radius of each individual as a measure of intraspecific density. Values for plants close to plot edges were divided by the proportion of the circle falling inside the plot.
In 2004, the slope, soil depth, litter depth, and density of interspecific plants were recorded. The aspect of the slope was recorded as N, S, E, or W, and inclination was measured with a ruler and a protractor. We assumed that the interaction of inclination and aspect mainly describes relative "warmth" of microsites. Accordingly, in linear models containing the interaction of these two variables, the effect of steeper north-facing slopes was a later start of flowering while for the other aspects, steeper slope was correlated with an earlier flowering. Thus, the slope measure was calculated by multiplying inclination values with 1.0 for south-facing slopes, 0.5 for east- and west-facing slopes, and –1.0 for north-facing slopes. Soil depth was measured with a 2 mm diameter, 35 cm long metal stick and was calculated as the mean of three measurements around each individual. Litter depth was measured next to each plant with a ruler. Interspecific density was estimated as the product of visual estimations of percentage cover and mean height of vegetation within 30 cm of the focus plant.
The remaining environmental variables were recorded in 2005. The canopy cover was photographed twice, in May before and in July after leaf development of deciduous trees. The camera was placed facing straight up on a 50 cm long stick next to each individual. Percentage canopy openness was calculated with the open source software ImageJ v. 1.34 (http://rsb.info.nih.gov/ij/). Eight soil samples were collected 5–15 cm deep in a circle of 0.5 m radius in each of the corners of all plots, with the center of the circle 1 m in from each plot edge. The eight samples per plot quadrant were pooled, and the resulting 100 samples were sent to a specialist lab (the Division of Soil Fertility and Plant Nutrition at the Swedish University of Agricultural Sciences) to determine concentrations of phosphorus, calcium, potassium, and magnesium (using the Al extraction method); soil pH, and total amounts of carbon and nitrogen. Additional soil samples at the same points were collected over 4 h on one day and immediately weighed. They were then dried at room temperature for 1 mo. Soil water content was measured by dividing the difference between the fresh and dry mass with the dry mass.
Temperature loggers, placed 10 cm below ground at the centers of the circles used for soil samples, recorded soil temperature every third hour for the period between shoot emergence and flower production, i.e., 6 May–26 May. Several temperature parameters were calculated per logger, including the mean of all temperature measurements, which corresponds to the temperature sum. The temperature parameters were intercorrelated, and we chose the one with the strongest correlation with flowering time for the analyses: the mean value of the highest temperature measurement of each day (hereafter, soil temperature). We did not measure air temperature because we expected differences in air temperature between sample points to be similar to differences in soil temperature.
The sampling procedure for soil characteristics yielded one observation per plot quadrant. Each Actaea individual was also given a unique value for soil nutrient concentrations, pH, water content, and temperature by calculating a weighted mean of the four values available per plot. Weighted values were calculated as sums of the four measured values multiplied by the inverse of the distances from the focal individual to the sampling points, divided by the sums of the inverse distances. Because continuous gradients of nutrient concentrations in the soil cannot be assumed (e.g., Lechowicz and Bell, 1991
), these values are only approximations. However, the values are conservative estimates because they are determined mostly by the closest measured value, but also by the plot mean.
Statistical analyses and model building
To determine variation in phenology at the population, plot, and individual level, variance components were calculated with the REML method in the lme4 package in the open source software R 2.4.1 (R Development Core Team, 2007
). Because of the small sample size (N = 4), causes of variation at the population level were not examined, and the rest of the analyses were performed at the plot and individual level. The difference between population and overall means was subtracted from plot means to remove the effect of population. The difference between plot and overall means was subtracted from individual values to remove the effect of plot. At both levels, as a screening of variables before multiple regression models were built, flowering time and shoot emergence were regressed on all other variables (Hosmer and Lemeshow, 2000
). At the individual level, nonsignificant predictor variables (P> 0.05) in the simple regressions were excluded from further analyses. At the plot level, the significance level for removal of variables was set at 0.01 to avoid having too many predictor variables relative to the sample size (N = 25). The predictor variables were shoot emergence, plant size, fruit number in the previous year, soil temperature, slope, soil depth, litter depth, interspecific density, intraspecific density, May canopy openness, July canopy openness, soil water content, soil pH, and soil concentrations of calcium, potassium, magnesium, phosphorus, carbon, and nitrogen. Because of the problems associated with having too many variables in the statistical modeling procedure, no ratios between concentrations of nutrients were used as variables.
The Akaike information criterion (AIC) of multiple regression models with all combinations of selected predictor variables was calculated in the wle package in R. Models with the lowest AIC containing only variables significant at P< 0.05 were chosen as final multiple regression models (Crawley, 2002
). At the plot level, no interactions between predictor variables were included. At the individual level, where sample sizes were larger, models without interactions between predictor variables and models including interactions between pairs of predictor variables were tested separately. Hierarchical partitioning of the fit of predictor variables in the final multiple regression models was carried out in the hier.part package in R. Significance of the soil characteristic variables was tested by fitting final multiple regression models using plot quadrant data (N = 75), which consists of completely independent measures of all variables.
Path models based on the results of the multiple regressions were constructed to study indirect effects on flowering time acting through shoot emergence. Structural equation modeling (SEM) was used to test whether the path models were rejected statistically (Shipley, 2000
). Interactions were added between predictor variables if they were significantly correlated in single tests. All paths between variables must be specified because tests of the fit of SEMs are made on the difference of the observed and predicted covariation matrices of all variables (Grace, 2006
). The directions of the added paths were specified if only one direction was plausible (i.e., slope could affect temperature, but not the other way around). Significance levels of individual paths were determined with t tests, and paths with P > 0.05 were removed. SEMs were also constructed by using the software Tetrad 3 (http://www.phil.cmu.edu/projects/tetrad/tet3/master.htm) to determine which variables were causally connected. The results were similar and are not presented. The SEMs were tested with maximum likelihood chi-square tests in the sem package in R. In accordance with common practice in structural equation modeling, models were accepted if p > 0.05 of the model being correct (Shipley, 2000
; Grace, 2006
). All statistical analyses were carried out in R 2.2.1 or R 2.4.1.
RESULTS
Approximately 51% of the variation in flowering time of Actaea spicata individuals occurred between populations, and 8% of the variation occurred between plots (variance components: population 8.26, interaction of population and plot 1.36, residual 6.58). Around 2% of the variation in shoot emergence occurred between populations and 24% between plots (variance components: population 0.0005, interaction of population and plot 0.0061, residual 0.0191). This leaves 41% of the variation in flowering time and 74% of the variation in shoot emergence for differences between individuals within plots.
Among-plot variation
Environmental variables were not strongly intercorrelated, except for nutrient concentrations (not shown). The best-fitting multiple regression model of early shoot emergence included temperature and July canopy openness and explained 63% of the variation (R2 = 0.63, P < 0.001) (Table 1). Early emergence was associated with high temperature and low canopy openness. Early flowering was related to early shoot emergence, steep and south-facing slopes, high temperature, and low July canopy openness. The final multiple regression model, including the environmental variables, explained almost all of the among plot variation in flowering time (R2 = 0.83, P < 0.001). Shoot emergence was tightly correlated with flowering time (R2= 0.64, P < 0.001), but including it as a predictor did not improve the flowering time model based on environmental factors only. The path model based on the results of the multiple regressions corroborated the result regarding the relationship between shoot emergence and flowering time, but was significantly improved by adding a connection between slope and shoot emergence (Fig. 1).
|
|
|
|
We found that within-population variation in flowering time of A. spicata was partly caused by differences in microenvironment. Within plots, small individuals on steeper, more south-facing slopes and on soils richer in potassium emerged earlier in the season and then flowered earlier. Among plots, flowering time was influenced by soil temperature, slope, and July canopy cover. These results support the findings that factors other than changes in temperature, photoperiod, and moisture affect flowering phenology. The indirect effect of individual environment on flowering time through shoot emergence indicates that evolutionary explanations of flowering phenology should include factors influencing vegetative phenology.
Differences in shoot emergence and flowering time were more strongly correlated with environmental properties among plots than within plots, even though most of the variation occurred within plots. At the within-plot level, only around 10% of the variation in flowering time was explained, whereas at the among-plot level, 83% of the variation was explained. The unexplained variance within plots may be due to any combination of genetic differences, unmeasured small-scale environmental heterogeneity, and individual history (McMillan and Pagel, 1958
; Wang, 1960
).
Earlier studies have found the "temperature sum" to fairly accurately predict differences in flowering time between years for many different species (e.g., Diekmann, 1996
). In accordance, a higher soil temperature was the most important cause of earlier shoot emergence and flowering time in plots in A. spicata. We also assume that the effects of slope and July canopy cover were partly due to temperature. However, the fact that there is an effect of slope on among-plot variation when we account for temperature differences suggests that the slope relates also to some unmeasured property of the environment. The association between low July canopy cover and early flowering at the among-plot level could be caused by cooler night air temperatures, which may not be reflected in the soil temperature (cf. Wielgolaski, 2001
). Alternatively, because A. spicata is adapted to shady conditions, dehydration may stress the plant, altering its phenology. Taken together, our results show that differences in temperature account for a large part of the within-population variation in flowering time of A. spicata.
Differences in microclimate are known to cause variation in flowering time (e.g., Jackson, 1966
; Marquis, 1988
), but environmental properties other than temperature, photoperiod, and moisture have not received much attention. In this study, high potassium concentration was associated with earlier shoot emergence at the within-plot level. There was also a marginally significant relationship between higher phosphorous concentration and earlier flowering. Because of intercorrelations of soil nutrient concentrations, we are not able to separate the effects of specific nutrients from a high general nutrient concentration. However, potassium has been shown to be a limiting nutrient of A. spicata in other studies (Salomonson et al., 1992
). Soil nutrients did not influence variation in phenology among plots, indicating that the largest variation of nutrient concentrations occurred at small scales (cf. Lechowicz and Bell, 1991
). Lack of insolation did not seem to delay shoot emergence or flowering time as later development was not associated with a more dense tree canopy, neither before nor after leaf development in deciduous trees. We also found no negative effects of intra- or interspecific densities on early flowering that would indicate effects of competition. Instead, high interspecific density and early shoot emergence were weakly and positively correlated at the within-plot level. This may have been caused by a clustering of individuals at spots with favorable, unmeasured aspects of the microenvironment.
The results show that, within a population of A. spicata, environmental conditions that can arguably be interpreted as favorable for this species are associated with earlier shoot emergence and flowering time. If the plants respond directly to the environment, then the effects of soil nutrients in this study indicate that seasonal timing of plant development is a more fine-tuned process than is often assumed. In accordance, nutrient concentrations are known to directly influence the phytohormone balances and development of plants (Marschner, 1986
). However, not all of the correlated environmental factors may act as cues of shoot emergence and flowering time. Phenology may be associated with environmental factors if plants must reach a certain resource state before initiating development.
If variation in phenology depends on the resource state of the plant, we would expect earlier shoot emergence and flowering time to be associated with larger plant size and less reproductive effort in previous years (Wang, 1960
; Schemske et al., 1978
). Plant size did influence shoot emergence in our study, but smaller plants emerged earlier than larger. This may simply result from larger shoots taking longer to develop than shorter shoots. Alternatively, it could reflect a more urgent need to compete for resources in small plants. A fitness gain through earlier emergence has been documented for seedlings in several studies (e.g., Verdu and Traveset, 2005
). That the number of fruits produced in 2004 was not correlated with shoot emergence or flowering time in 2005 suggests a lack of effect of reproductive effort the previous year.
Concluding remarks
In addition to temperature, the slope, canopy openness, and soil nutrient concentrations influenced plant phenology. Important factors differed over spatial scales and influenced vegetative phenology, flowering phenology, or both. Plants that emerged earlier also flowered earlier. The abiotic factors influencing phenology may be a source of spatial variation in plant–animal interactions over different spatial scales because vegetative and flowering phenology are important for biotic interactions. The results also indicate that soil nutrient concentrations should be considered when predicting how climate change will influence plant phenology.
|
1 The authors thank S. Rosengren, T. Lund Jörgensen, and M. Bengtsson for assistance with the field work. This study was supported financially by grants from the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (to J.E.). 
2 Author for correspondence (e-mail: johan.dahlgren{at}botan.su.se
)
LITERATURE CITED
Augspurger C. K.. 1981. Reproductive synchrony of a tropical shrub: experimental studies on effects of pollinators and seed predators on Hybanthus prunifolius (Violaceae). Ecology 62: 775-778..[CrossRef][Web of Science]
Beaubien E. G. Hall-Beyer M.. 2003. Plant phenology in western Canada: trends and links to the view from space. Environmental Monitoring and Assessment 88: 419-429..[CrossRef][Web of Science][Medline]
Brody A. K.. 1997. Effects of pollinators, herbivores, and seed predators on flowering phenology. Ecology 78: 1624-1631..[Web of Science]
Crawley M. J.. 2002. Statistical computing, an introduction to data analyses using S-Plus. Wiley, Chichester, UK..
Debussche M. Garnier E. Thompson J. D.. 2004. Exploring the causes of variation in phenology and morphology in Mediterranean geophytes: a genus-wide study of Cyclamen. Botanical Journal of the Linnean Society 145: 469-484..[CrossRef][Web of Science]
Diekmann M.. 1996. Relationship between flowering phenology of perennial herbs and meteorological data in deciduous forests of Sweden. Canadian Journal of Botany 74: 528-537..
Diggle P. K.. 1999. Heteroblasty and the evolution of flowering phenologies. International Journal of Plant Sciences 160: S123-S134..[CrossRef][Web of Science][Medline]
Eriksson O.. 1994. Seedling recruitment in the perennial herb Actaea spicata L. Flora 189: 187-191..
Eriksson O.. 1995. Asynchronous flowering reduces seed predation in the perennial forest herb Actaea spicata. Acta Oecologia 16: 195-203..
Evans E. W. Smith C. C. Gendron R. P.. 1989. Timing of reproduction in a prairie legume: seasonal impacts of insects consuming flowers and seeds. Oecologia 78: 220-230..[CrossRef][Web of Science]
Fitter A. H. Fitter R. S. R.. 2002. Rapid changes in flowering time in British plants. Science 296: 1689-1691..
Givnish T. J.. 1987. Comparative studies of leaf form: assessing the relative roles of selection pressures and phylogenetic constraints. New Phytologist 106: 131-160..[Web of Science]
Goulart M. F. Filho J. P. L. Lovato M. B.. 2005. Phenological variation within and among populations of Plathymenia reticulata in Brazilian Cerrado, the Atlantic forest and transitional sites. Annals of Botany 96: 445-455..
Grace J. B.. 2006. Structural equation modeling and natural systems. University Press, Cambridge, UK..
Hosmer D. W. Lemeshow S.. 2000. Applied logistic regression. Wiley, New York, New York, USA..
Hultén E. Fries M.. 1986. Atlas of north European vascular plants, vol. 1. Koeltz Scientific Books, Königstein, Germany..
Inouye D. W. Saavedra F. Lee-Yang W.. 2003. Environmental influences on the phenology and abundance of flowering by Androsace septentrionalis (Primulaceae). American Journal of Botany 90: 905-910..
Jackson M. T.. 1966. Effects of microclimate on spring flowering phenology. Ecology 47: 407-415..[CrossRef][Web of Science]
Lechowicz M. J. Bell G.. 1991. The ecology and genetics of fitness in forest plants. II. Microspatial heterogeneity of the edaphic environment. Journal of Ecology 79: 687-696..[CrossRef][Web of Science]
Lieth H., editor. 1974. Phenology and seasonality modeling. Springer-Verlag, New York, New York, USA..
Mahoro S.. 2002. Individual flowering schedule, fruit set, and flower and seed predation in Vaccinium hirtum Thunb. (Ericaceae). Canadian Journal of Botany 80: 82-92..
Marquis R. J.. 1988. Phenological variation in the neotropical understory shrub Piper arieianum: causes and consequences. Ecology 69: 1552-1565..[CrossRef][Web of Science]
Marschner H.. 1986. Mineral nutrition in higher plants. Academic Press, London, UK..
McMillan C. Pagel B. F.. 1958. Phenological variation within a population of Symphoricarpos occidentalis. Ecology 39: 766-770..[Medline]
Ollerton J. Lack A. J.. 1992. Flowering phenology—an example of relaxation of natural selection. Trends in Ecology and Evolution 7: 274-276..[CrossRef]
Pellmyr O.. 1984. The pollination ecology of Actaea spicata (Ranunculaceae). Nordic Journal of Botany 4: 443-456..[Web of Science]
R Development Core Team.. 2007. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07–0, URL http://www.R-project.org..
Rathcke B. Lacey E. P.. 1985. Phenological patterns of terrestrial plants. Annual Review of Ecology and Systematics 16: 179-214..[CrossRef][Web of Science]
Salomonson A. Ohlson M. Ericson L.. 1992. The effect of potassium on growth and nutrient uptake in two forest herbs with different chemical defence systems. Oikos 65: 493-501..[CrossRef][Web of Science]
Schemske D. W. Willson M. F. Melampy M. N. Miller L. J. Verner L. Schemske K. M. Best L. B.. 1978. Flowering ecology of some spring woodland herbs. Ecology 59: 351-366..[CrossRef][Web of Science]
Schnelle F.. 1955. Pflanzenphänologie. Geest & Portig, Leipzig, Germany..
Shipley B.. 2000. Cause and correlation in biology. University Press, Cambridge, UK..
Sola A. J. Ehrlén J.. 2007. Vegetative phenology constrains the onset of flowering in the perennial herb Lathyrus vernus. Journal of Ecology 95: 208-216..[CrossRef][Web of Science]
Verdu M. Traveset A.. 2005. Early emergence enhances plant fitness: a phylogenetically controlled meta-analysis. Ecology 86: 1385-1394..[CrossRef][Web of Science]
von Zeipel H. Eriksson O. Ehrlén J.. 2006. Host plant population size determines cascading effects in a plant–herbivore–parasitoid system. Basic and Applied Ecology 7: 191-200..[CrossRef][Web of Science]
Wang J. Y.. 1960. A critique to the heat unit approach to plant response studies. Ecology 41: 785-790..[CrossRef][Web of Science]
Wielgolaski F. E.. 2001. Phenological modifications in plants by various edaphic factors. International Journal of Biometeorology 45: 196-202..[CrossRef][Web of Science][Medline]
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Milla, L. Gimenez-Benavides, and G. Montserrat-Marti Replacement of Species Along Altitude Gradients: The Role of Branch Architecture Ann. Bot., December 1, 2008; 102(6): 953 - 966. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Bustamante and A. Burquez Effects of Plant Size and Weather on the Flowering Phenology of the Organ Pipe Cactus (Stenocereus thurberi) Ann. Bot., December 1, 2008; 102(6): 1019 - 1030. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Compiled by, F. Tooke, T. Chiurugwi, and N. Battey Flowering Newsletter bibliography for 2007 J. Exp. Bot., July 18, 2008; (2008) ern109v1. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
2 = 6.01, df = 3, p = 0.11).