T. J. De Jong

On the analysis of size-dependent reproductive output in plants

A general model is presented to analyse size-dependent reproductive output in plants. The model formulation makes it possible to test for both a minimum size for reproduction and a non-linear relationship between size and reproductive output. The model is statistically analysed with the likelihood-ratio test. In contrast to previously proposed models, this model admits a hump-shaped relationship between reproductive effort (RE) and plant size. Data for three species analysed here show that RE is not always monotonically increasing or decreasing with plant size (...)

Population ecology of the biennials Cirsium vulgare and Cynoglossum officinale in a coastal sand-dune area

(1) Demographic data were collected on the biennials Cirsium vulgare and Cynoglossum officinale in a coastal sand-dune area in the Netherlands. Over ninety local populations were observed in 2 m x 2 m plots for a three-year period (1981-84). (2) Within two years Cirsium vulgare disappeared from 51% of the plots, while Cynoglossum officinale disappeared from only 7% of the plots. Reduction of the total dune population during the study period was evident for both species; the number of seedlings that established was insufficient to compensate for the losses from the population because of death of rosettes and flowering plants. (3) Large temporal and spatial variation existed in reproductive success, measured as the number of surviving seedlings per reproductive individual. Seedling survival in the two species was lowest in the dry summer of 1983 and in exposed vegetation. As a result of the low establishment in open habitats the main distribution shifted from exposed habitats to plots in the direct vicinity of shrubs and trees. This shift was most obvious in Cynoglossum officinale. (4) Using a modified key factor analysis, plots with high early mortality also showed high total mortality during the life of the 1981 cohort: events early in the life of individuals determine the spatial differences in abundance of the adults. These events include failure to produce viable seeds, seed predation and early seedling mortality in Cirsium vulgare, and failure to produce viable seeds, failure to germinate and early seedling mortality in Cynoglossum officinale. (5) The reduction of the population appeared not to be due to systematic processes such as the accumulation of litter or the increase of perennial vegetation. Stochastic factors such as rainfall and seed predation had most influence on population recruitment. (6) Dispersal, delayed germination and delay of flowering permit a monocarpic organism to bridge years or sites with failing reproduction. In these two species the importance of a persistent seedbank in the soil is limited. The dispersal rate between populations in different vegetation types was probably low. Delayed flowering was most pronounced in Cynoglossum officinale. The proportion of the rosettes of one year and older delaying reproduction varied between 75% and 95%. In Cirsium vulgare this figure varied between 53% and 80%. The large proportion of rosettes that delayed reproduction buffered the population against low seedling recruitment in some years.

The importance of small-scale disturbance for seedling establishment in Cirsium vulgare and Cynoglossum officinale.

(1) Germination and establishment of Cirsium vulgare and Cynoglossum officinale in various micro-environments were studied in a sand-dune area at Meijendel, The Netherlands. (2) The number of established seedlings was increased forty-fold in Cirsium vulgare and five-fold in Cynoglossum officinale by artificial speed input of 11 seeds dm-2. In Cirsium vulgare, seed input was the primary prerequisite for seedling emergence. Disturbance of soil and vegetation alone had little effect on the number of seedlings establishing naturally, but increased the number of seedlings in plots with seeds sown. (3) Disturbance had the additional effect of advancing the germination period and accentuating autumn germination in Cirsium vulgare. (4) Seedling mortality did not differ between disturbed and undisturbed plots. In both species seedling recruitment was high in thickets and Crateagus scrub and low in open vegetation. (5) After seed sowing, significantly less seedlings were found in May and July on sites where the naturally occurring population had disappeared (Cirsium vulgare) or dropped to a low density (Cynoglossum officinale) than on sites where the natural population still thrived. These differences were not found (Cynoglossum officinale) or reduced (Cirsium vulgare) on disturbed plots with seeds sown, indicating the importance of small-scale disturbances for population maintenance.

Seedling establishment of the biennials Cirsium vulgare and Cynoglossum officinale in a sand-dune area: the importance of water for differential survival and growth

(1) Seedling mortality of the biennials Cirsium vulgare and Cynoglossum officinale was studied in a sand-dune area at Meijendel, The Netherlands. In both species seedling survival was positively correlated with the water content of the top 10 cm of the soil. Early seedling survival (April-May) in Cynoglossum officinale was also positively correlated with the cover of shrubs and trees. Soil water content and cover of woody plants explained 74% of the variance in the early mortality of Cynoglossum seedlings. (2) Throughout most of the year the water content of the top 10 cm of the soil was highest in thicket and lowest in open vegetation. Below 10 cm the rank order of the vegetation types with respect to soil water content was not consistent, and the total soil water potential was always above the critical level for water uptake by the plant. (3) Artificial watering increased seedling survival and growth in the two species, especially in open vegetation. Vegetation types differed with respect to seedling survival but these differences were greatly reduced by the addition of water. It is likely that variation in rainfall causes yearly differences in survival and growth in the two species and affects their distribution over shaded and exposed habitats.

Production, dispersal and predation of seeds in the biennial Cirsium vulgare.

(1) Seed production of flowering plants of Cirsium vulgare without damage to the flowering stem varied strongly between years in the sand-dunes at Meijendel. The percentage of plants that were noticeably damaged by herbivores also varied between years, and caused a reduction of the seed production by 80% in 1982 and 15% in 1984. (2) Mean seed loss by predation in the flowerheads was 3% in 1982 and 17% in 1984. (3) Seed dispersal was limited: 50% of the seeds were dispersed within 1 m of the parent plant and only 11 % of the seeds were dispersed outside the local population. (4) On average 60% of the seeds were consumed by mice and voles after dispersal. Only 1% of the seeds produced were still viable after one year.

Delay of Flowering in the 'Biennial' Cirsium vulgare: Size Effects and Devernalization

Delay of flowering is a common phenomenon in natural populations of Cirsium vulgare. In all years more than 45% of the rosettes that were at least one year old delayed reproduction. The probability of flowering increased with (in order of importance) increasing size, growth over the period immediately prior to bolting, and age. In contrast with the general idea about biennials that only large plants are vernalized during winter, we show that in C. vulgare all individuals can be vernalized in winter. Rosettes, from the smallest size-classes, collected from the field in Februari showed 80% flowering when placed under optimal conditions. We argue that in C. vulgare no size requirement exists at the time of vernalization but that flowering is delayed if plants fail to attain a required minimum size for bolting (probably induced by long days) in May.

Glucosinolates and other metabolites in the leaves of Arabidopsis thaliana from natural populations and their effects on a generalist and a specialist herbivore

Abstract.Because many secondary metabolites in plants act as defense against herbivores it has been postulated that these compounds have evolved under selective pressure by insect herbivores. One explanation for the within-species variation in metabolite patterns in a particular species is that different populations are under selection by different herbivores. We tested this hypothesis, using Arabidopsis thaliana plants that originated from dune and inland areas. We analyzed Arabidopsis thaliana leaves using NMR spectroscopy and multivariate data analysis. Major differences in chemical composition were found in water-methanol fractions and were due to higher concentrations of sinigrin and fumaric acid in dune plants. Inland plants showed lower levels of glucose. Quantitative analysis of glucosinolates was performed with HPLC. Individual plants and populations demonstrated differences in glucosinolate composition and concentration. In growth chamber experiments, the generalist herbivore, Spodoptera exigua grew significantly better on the inland plants, while the specialist herbivore Plutella xylostella performed equally well on plants of both origins. Aliphatic glucosinolate as well as total glucosinolate concentrations negatively correlated with larval mass of Spodoptera exigua. No significant correlations, however, were found between larval mass of Plutella xylostella and glucosinolates in the leaves. A specialist and a generalist herbivore were responding differently to plant secondary chemistry, as was also found in several other studies. This is an important indication that differences in glucosinolate concentrations among populations may result from differential selection by different guilds of herbivores.

The Control of Flowering in the Monocarpic Perennial Carlina vulgaris

In natural populations of the biennial Carlina vulgaris more than 95% of the individuals delayed flowering past the second year. The probability of flowering increased with (in order of importance) increasing size before boltining in June, size before the winter (November of the year before flowering) and growth in spring (April to June). No significant effect of age was found. In a nutrient rich experimental garden 58% of the individuals flowered in their first year. Laboratory experiments showed that all individuals flowered without cold if they were given optimal growing conditions (...)

Effects of nectar content on the number of bumblebee approaches and the length of visitation sequences in Echium vulgare (Boraginaceae)

The effect of nectar reward on a bumblebees decision to stay on or to leave an individual plant of Echium vulgare was studied in the field. Nectar standing crop was varied by covering plants for different periods. A stochastic model is developed which describes the bumblebees decision strategy as a function of the duration (in seconds) of the current flower visit (y The duration of the previous flower visit and the number of previous flower visits on the same plant did not influence the bumblebees departure behaviour. Bumblebees on E. vulgare use a probabilistic decision rule. The chance (p) of staying on the plant can be described by a hyperbolic function of the time spent in the current flower (y ): p = 0.97 - (0.11/(0.04 + y)). Simulations were carried out in which the distribution of nectar over the flowers within a plant was varied. High variation in nectar content shortened visitation sequences. In the field we covered plants to increase nectar standing crop per flower. This resulted in longer visitation sequences, but no effect of nectar content on the number of approaches was found. This is attributed to the fact that the first few bumblebees which approach a plant empty most flowers within a relatively short period with respect to the total observation time.

Metabolic fingerprinting reveals differences between shoots of wild and cultivated carrot (Daucus carota L.) and suggests maternal inheritance or wild trait dominance in hybrids.

Differences between the metabolic content of cultivars and their related wild species not only have implications for breeding and food quality, but also for the increasingly studied area of crop to wild introgression. Wild and cultivated western carrots belong to the same outcrossing species and hybridize under natural conditions. The metabolic fingerprinting of Dutch wild carrot and of western orange carrot cultivar shoots using (1)H NMR showed only quantitative differences in chemical content, indicating relatively low divergence after domestication. Main differences reside in the primary metabolite content and in the concentrations of chlorogenic acid and feruloyl quinic acid in the shoots of the different carrot types. Wild×cultivar hybrids cannot be distinguished from wild plants based on the metabolome, suggesting maternal, maternal environment, or dominance effects, and indicating high hybrid fitness in wild conditions. Considering these similarities, introgression is a real possibility in carrots, but understanding its consequences would require further studies using backcrosses in a multiple environments.