Search

Jebel Marra, a volcanic complex in western Sudan, is an important site of early settled agriculture, with high plant diversity, supported by orographic rainfall. Vegetation types were examined in r...

This study was designed to examine the subtle effects of habitat, aspect and altitude on plant diversity and the timing of phenological stages. In this increasingly arid region, within-species differences would cause more reproductive isolation of local populations, making species more vulnerable to regional extinction. Fifty-two sites in the Jebel Marra region of the Central Darfur state of Sudan were compared in terms of species diversity and local variation in the phenology of plant communities. Five habitat types were compared: sandy plains, clay plains, wadis, north-facing slopes and south-facing slopes. Sites were classified into five altitudinal zones ranging from 750 m to > 1,150 m above sea level, and the phenology of three categories of plant life form (phanerophytes, chamaephytes and therophytes) was tracked in terms of the duration of each phenophase (vegetative, reproductive and dormant). Overall, most plants were dormant from December to April. Peak vegetative growth was in July, coinciding with peak rainfall in all study sites in the region (July-August). This was followed by peak reproduction in October. Life form and habitat type significantly affected the length of all phenological phases; altitude significantly affected all stages except the duration of dormancy. Some of the interactions between these factors were also statistically significant, indicating subtle regulation of the duration of phenophases. The possible impacts of overgrazing and ongoing climate change are discussed, with the concern that initiation of vegetative growth in most habitats (except the clay plains) appears to be triggered by day length, ‘anticipating’ the beginning of the rainy season. Further climate change may uncouple this signal and the beginning of the rainy season, impacting plant survival.

Background: The lives of the “Fur”, indigenous people of Darfur, Sudan are intimately connected to local wild plants, but the traditional uses of these plants are, so far, poorly documented. Many species are indigenous to the region, but others are introduced, and have naturalized over millennia. Methods: For a month each summer from 2014-2016, using questionnaire interviews, direct observation of practices, and a literature review, 58 species were identified. An “importance value” for each was determined, based on the intensity and season(s) of use, in eight use categories. For each species, a “concordance ratio” characterized the degree of agreement between indigenous knowledge and our current “scientific understanding” of their value. Results: All species were multi-use; animal forage, “other functional uses”, traditional medicine, and construction predominated. Some species are declining due to overharvesting by the growing local population, exacerbated by conflict and refugee encampments. Most of the species are used in traditional medicines, but active ingredients have been scientifically confirmed for only half of them. Surprisingly, several species with known medicinal ingredients are not used locally. Conclusions: The “Fur” people have long combined agriculture with pastoralism and wildcrafting. For this to be sustainable, it is critical to understand cultural contexts and recognize multi-use species. This can help identify new medicines, and guide sustainable development of local resources, adapted to local conditions. Naturalized wild fruit trees may have evolved drought resistance in this increasingly dry savanna climate; such genes might usefully be incorporated in crop strains elsewhere as climate change proceeds. Key words: Ethnobotany, Multipurpose-Use Plants, Importance Value, Indigenous Knowledge Index, Scientific Knowledge Index, Concordance Value, Plant Diversity, Climate Change

Within the semi-desert landscape of northern Libya, two sub-humid escarpments occur: Al-Akhdar in the east and Nafusa (Jabal Al-Gharbi) in the west. This study compares plant communities in the two regions, which are along an elevation gradient, in terms of species composition and diversity, frequency of different Raunkiaer life forms, and reproductive phenology. The two regions differed in species composition and life-form frequency between regions and between elevation zones within each region. Patterns were associated with the lower rainfall and lower moisture-holding capacity of soils at Nafusa, resulting in more xeric conditions. Only 13% of species were shared between the two regional landscapes. Species diversity, life-form frequency, and duration of the flowering–fruiting phenophase were all affected by elevation above sea level. The duration of flowering and fruiting in spring and fall was associated with environmental conditions, although there were different thresholds in the two regions. There was both a spring and fall episode of flowering at Nafusa, but only spring flowering at Al-Akhdar. It is anticipated that there will be a gradual shift of plant communities to higher elevations and loss of certain sensitive species in response to ongoing climate change.

We sampled twenty populations of the vulnerable endemic shrub or tree, Arbutus pavarii Pampan., at different elevations and aspects within the Al-Akhdar mountainous region of Libya. Our sampling sites were at elevations ranging from 285 to 738 m above sea level, and several different habitats: vallies (locally known as wadis), north- and south-facing slopes, and mountaintops. All individuals within each quadrat were studied. Population size and structure, and plant functional traits were assessed. None of the populations had a stable distribution of size classes. Some consisted mostly of small plants, with little or no fruit production; others consisted only of mid-sized and large plants, with high fruit production, but no juvenile recruitment. There was a significant increase in percent cover with increasing elevation; reproductive output (the number of fruits per branch and total number of fruits per individual) also generally increased with elevation. In some of these populations the lack of recruitment, and in others the failure to produce fruit, together constituted serious demographic threats. In light of these results, recommendations are made for conservation of this vulnerable endemic species.

The Phoenician juniper (Juniperus phoenicea L.) has long been common in the Al-Akhdar mountain range in NE Libya. Juniper trees in the region are showing signs of dieback, and minimal successful seedling recruitment has occurred. We investigate this effect at the level of individual trees, tree size classes, and populations. At ten sites, population size, growth and reproductive traits, and percentage dieback in different size classes were evaluated. Elevation, distance from the sea, and the interaction between these factors had a significant effect on tree volume, and number of branches, number of female cones per branch and number of female cones per tree. Two groups of populations were recognized: the first (Sites 2, 3, and 4) were near the coast. They appeared to be more recently established, with no individuals in the largest size class, and, at this point, less severe dieback. In contrast, in the rest of the sites there were no individuals (Sites 1, 5, 6, and 9) or very few individuals (Site...

The cost of reproduction by female plants generally exceeds that of male plants because females produce seeds and fruits in addition to flowers (Darwin 1877, Lloyd and Webb 1977, Willson 1983). If the vegetative growth of female and male plants is equivalent, then the greater female reproductive effort should result in reduced growth and/or survival. In a dioecious clonal plant that reduction might be manifest as lower ramet production, and the sex ratio of ramets should become increasingly male-biased. In contrast to this expectation the ramet sex ratio of Rumex acetosella tends to be female-biased (Harris 1968, Putwain and Harper 1972, Lovett Doust and Lovett Doust 1985), even though females allocate a greater proportion of their resources to reproduction than do males (Putwain and Harper 1972). In eight Massachusetts populations of Rumex acetosella, we found one male-biased population and a female bias at the other sites. In these latter seven sites the estimated sex ratio of genets (the genetic individuals) did not differ significantly from 1: 1 (Lovett Doust and Lovett Doust 1985). The greater reproductive effort by female plants, plus the female bias in sex ratio of ramets suggested differential growth by females. Biased sex ratios in clonal plants can be caused by a variety of factors, including, for example, differential grazing by herbivores (see Lovett Doust and Lovett Doust 1985, Agren 1987), sex-associated effects of intraspecific competition (e.g., Onyekwelu and Harper 1979, Lovett Doust et al. 1987), and differences in the numbers of ramets produced and sustained by female and male plants (Sakai and Burris 1985). We have examined and rejected the possibility that differential herbivory on male and female Rumex plants drives the ramet sex ratio to being female-biased. (In fact, although males and females suffered comparable herbivory in absolute terms, the proportion of the available foliage that was removed from females was greater [Lovett Doust and Lovett Doust 1985].) The present study was designed to explore the hypothesis, suggested by David Lloyd (personal communication), that differences in the vegetative vigor and seasonal phonology of males and females might explain the female-biased sex ratios generally found in populations of Rumex acetosella.

Leaf-to-root surface area ratios in Vallisneria americana have been shown to provide a simple and inexpensive relative measure of sublethal effects of organochlorine contamination. The present study was conducted to determine whether this index of surface area could be used as an effective biomonitor of overall site quality in stressed aquatic ecosystems. The leaf-to-root surface area ratio was determined for samples of V. americana collected from 225 microsites within 12 Areas of Concern (environmentally contaminated areas designated by the International Joint Commission) throughout the Laurentian Great Lakes of Ontario. Statistical analyses indicate that 77% of the variation in the surface area index could be attributed to differences among microsites, with only 23% of variation occurring among plants within a microsite. A multiple regression equation was developed for predicting the leaf-to-root surface area ratio from several measures of microsite quality. Significant parameters affecting the surface area ratio included plant density, light intensity, and an index of sediment contamination. In contrast, measures of water contamination did not show any correlation with leaf-to-root surface area ratio. These observations support the hypothesis that V. americana accumulates contaminants primarily from the sediments and that the leaf-to-root surface area ratio can be used to construct contours of point source impact zones in Areas of Concern. The regression model developed here provides a simple, inexpensive means for monitoring overall site quality throughout the Great Lakes.

We examined the development of ovules and components of reproductive yield as they related to plant size and fruit position within plants of Alliaria petiolata (M. Bieb.) Cavara and Grande from three sites. We found that both factors affected fruit and seed production, and patterns of ovule fate. Larger plants were more fecund than their smaller counterparts, while proximally positioned flowers within an inflorescence produced more seeds. Maternal investment was regulated according to plant mass at two levels: among fruits (small plants had significantly lower fruit-set than did large plants), and within fruits (small plants had significantly fewer ovules and significantly higher proportions of aborted ovules per fruit than did large plants). All absolute measures of reproduction had positive linear relationships with plant size, but a threshold size for flowering was not found. Most proportional measures of reproduction were independent of plant size. Small plants produced the same relative proportions of total seed number and total seed mass per unit plant dry mass as large plants. Therefore, proportional reproductive allocation did not vary with plant size. However, small plants produced proportionately more flowers per unit dry mass. Small plants may be maximizing their overall fitness through increased allocation to male function (flower production).Key words: Alliaria petiolata, plant-size effects, fruit-position effects, reproductive allocation, patterns of ovule abortion, seed maturation.

The clonal perennial herb, Arisaema dracontium (L.) Schott., is considered "vulnerable" in Canada but is more abundant in the adjacent northeastern United States. In Canada, sexual reproduction may be limited since the species is self-incompatible, and populations at the edge of a species' range may contain limited genetic diversity. We compared genetic variation in nine enzymes systems, sampled at 19 populations across eastern North America. Based on chromosome counts and interpretable banding patterns, A. dracontium is tetraploid over most of its range (2n = 56). However, plants in Florida were diploid, lacked 10 out of a total of 19 interpretable alleles, had distinct allelic frequencies, and clustered a large distance from the other populations. Some plants from Baton Rouge, Louisiana, were also diploid, but all allozyme samples were tetraploid. Most genetic variation occurred within populations; only 29% of variation was between populations. Individuals in two adjacent, otherwise isolated Ontario populations were fixed for most loci and were balanced heterozygotes for the got-2 locus. All samples but one were also identical for an additional six non-interpretable enzyme banding patterns, suggesting these two populations comprise multiple ramets of a single multilocus genotype, propagated clonally via corms. Ecologically marginal populations were typically more distant from the nearest neighboring population and were more genetically distinct from one another; however, with the exception of the two monoclonal populations, marginal populations were not genetically depauperate.

In four field and glasshouse experiments designed to alter the supply of resources through manipulation of nutrients, root tissue, leaf area and fruit number in Alliaria petiolata (Brassicaceae), more than 99% of ovules per plant showed signs of fertilization, suggesting that seed production in this plant was not pollen limited. However, in all treatments a significant proportion of fruits and seeds did not develop to maturity. Total fruit and seed production did not differ significantly from controls when plants were given nutrient supplements at flowering. Removal of 50–75% of the root tissue in 1-yr-old plants significantly reduced fruit set, but had no effect on individual seed development. Removal of cauline leaves significantly reduced most measures of fruit and seed production, suggesting that current photosynthate is critical for fruit and seed filling. Seed maturation was significantly affected by both fruit position within an infructescence and ovule position within a fruit. Basally located fruits and ovules (within fruits) developed more mature seeds than distally positioned fruits and ovules. Plants responded to removal of basal fruits by re-allocating resources to distal fruits that would normally have aborted. Our results suggest that fruits and seeds act as reproductive sinks competing for parental photosynthate. Patterns of resource allocation within infructescences and fruits were also modified by our experiments.

Arisaema dracontium (green dragon) is a perennial herb that is widely distributed in eastern North America. However, in Canada, at the northern edge of its distribution, the species is designated as "vulnerable" with respect to conservation status. In natural populations, seedlings are uncommon; the present study was undertaken in order to characterize seed and seedling properties in green dragon. Seeds were sampled from five sites, ranging from Ontario at the northern limit of the distribution range, to Louisiana in the south. Seed germinability ranged from 25 to 55%, depending upon source. Experiments indicated that neither the hard seed coat nor a water-soluble exudate from the seed was responsible for inducing or maintaining dormancy. Patterns of seed germination appear to reflect general climatic conditions at the sites where seeds had originated. Cold stratification at 3°C produced significantly greater relative germinability in all seed collections except the most southerly one, from Baton Rouge. These seeds also had a slower overall speed of germination. In contrast, germination of seeds from the most northerly site was promoted by cold stratification and occurred over a relatively brief period. Germination in alternating light and dark conditions decreased the speed of germination compared to germination in the dark, however exposure to light changed the phenology of germination by promoting development of adventitious roots and primary leaves in these seedlings.

Morph-specific patterns of success as maternal and paternal parents were studied in a factorial pollination experiment with long-, mid- and short-floral morphs in tristylous Lythrum salicaria L. `Legitimate' pollinations (involving stamens and stigma of similar heights) produced greater fruit and seed set than `illegitimate' intermorph pollinations (involving stamens and stigma of dissimilar heights between different morphs), `illegitimate' intramorph pollinations (between two individuals of the same floral morph) and self-pollinations (pollination involving stamens and stigma of the same flower). The mean rate of germination of seeds produced by legitimate and illegitimate crosses was 81.4% and 72.7%, respectively. The three flower morphs differed significantly in their ability to act as maternal parent but did not differ paternally, following legitimate pollinations. Significantly greater legitimate seed set was recorded for the long morph compared to the short morph, when these individuals acted as maternal parents. Incompatibility (as measured by seed set) was weakest in the mid-morph individuals and strongest in short-morph individuals. Following self-pollination events, pollen from the mid- and long whorls of stamens within a flower produced seeds in the long and mid-morphs, respectively. This suggests that the self-compatibility differs between the two whorls of stamens, and is greater in the longer stamens than in the shorter stamens in the long and mid-morphs.

We investigated variation in ovule development within and among fruits in garlic mustard, Alliaria petiolata (M. Bieb.) Cavara & Grande. Individuals were sampled at 14 sites in North America. On average, 94% of ovules showed evidence of fertilization and development. The majority of ovules (mean 68%; range 53.2-82.5%) reached seed maturity. Patterns of seed maturation and abortion varied nonrandomly within and among fruits. In sites in Tennessee, Kentucky, and Ontario, resource limitation seemed to be an important determinant of seed production, as fruits initiated nearest the base of an infructescence were significantly more likely to produce mature seeds than distally located fruits. Also the probability of seed maturation within individual fruits decreased significantly from the pedicellar to the stylar ends. In contrast, for individuals from Ohio sites, the probability of maturing seeds was greatest for centrally located fruits within an infructescence as well as for centrally located ovules within fruits, indicating a greater influence of pollen limitation in addition to resource limitation. We conclude that it could be misleading to infer traits of a species as a whole based on observed patterns of seed maturation in a single site or region. Rather such patterns may reflect prevailing conditions and selection pressures at a local, or regional, scale.Key words: Alliaria petiolata, Brassicaceae, garlic mustard, ovule abortion, seed maturation, patterns.

We investigated variation in ovule development within and among fruits in garlic mustard, Alliaria petiolata (M. Bieb.) Cavara & Grande. Individuals were sampled at 14 sites in North America. On average, 94% of ovules showed evidence of fertilization and development. The majority of ovules (mean 68%; range 53.2-82.5%) reached seed maturity. Patterns of seed maturation and abortion varied nonrandomly within and among fruits. In sites in Tennessee, Kentucky, and Ontario, resource limitation seemed to be an important determinant of seed production, as fruits initiated nearest the base of an infructescence were significantly more likely to produce mature seeds than distally located fruits. Also the probability of seed maturation within individual fruits decreased significantly from the pedicellar to the stylar ends. In contrast, for individuals from Ohio sites, the probability of maturing seeds was greatest for centrally located fruits within an infructescence as well as for centrally located ovules within fruits, indicating a greater influence of pollen limitation in addition to resource limitation. We conclude that it could be misleading to infer traits of a species as a whole based on observed patterns of seed maturation in a single site or region. Rather such patterns may reflect prevailing conditions and selection pressures at a local, or regional, scale. Resume : Les auteurs ont examine la variation du developpement de l'ovule a l'interieur et entre les fruits de l'Alliaria petiolata (M. Bieb.) Cavara & Grande (ail-moutarde). Ils ont echantillonne des individus provenant de 14 sites en Amerique du Nord. Globalement, 94% des ovules montraient des signes de fecondation et de developpement. La majorite des ovules ont atteint (en moyenne 68%; ecart 53,2 a 82,5%) la maturite. Les patrons de maturation des graines et d'avortement varient de facon non-aleatoire a l'interieur et entre les fruits. Sur les sites du Tennessee, du Kentucky et de l'Ontario, la limitation en ressources semble etre un important facteur determinant pour la production des graines, puisque les fruits se formant tout pres de la base de l'infructescence sont significativement plus enclins a produire des graines mures que les fruits se formant en position distale. De plus, la probabilite de maturation des graines a l'interieur du fruit individuel diminue significativement de l'extremite pedicellaire en allant vers celle du style. Au contraire, chez les individus provenant des sites de l'Ohio, la probabilite que les graines murissent est plus elevee pour les fruits localises au milieu, a l'interieur de l'infructescence, aussi bien que pour les ovules localisees centralement a l'interieur du fruit, ce qui indique une plus grande influence d'une pollinisation limitee en plus de la limitation en ressources. Les auteurs concluent qu'il pourrait etre trompeur de deduire les caracteres d'une espece dans son ensemble en ne se basant sur les patrons de maturation des graines, a partir d'une seule station ou region. De tels patrons peuvent au contraire refleter les conditions et les pressions selective prevalant a l'echelle locale ou regionale. Mots cles : Alliaria petiolata, Brassicaceae, ail-moutarde, avortement ovulaire, maturation des graines, patrons. (Traduit par la Redaction) Susko and Lovett-Doust 1686

Macrophytes drifting in the Detroit River were sampled and analysed for contaminants at monthly intervals from September 1990 to September 1991. Twelve species of submersed macrophytes were identified, as well as algae and leaves of terrestrial plants. Drifting plant debris was greatest in August-September, when Potamogeton spp. and Najas sp. predominated. Over the study period, a total of 60.57 times 106 kg fresh mass (3.0285 times 106 kg ash-free dry mass) of plant debris drifted out of Lake St. Clair into the Detroit River annually. Organochlorine content differed between plant taxa and according to the time of year. Annual contaminant burden of the Detroit River by upriver contributions was carried mostly by Potamogeton spp. and Najas sp. Total annual load of organochlorines in drifting plant debris was estimated to be 155 g, including 124 g of PCBs. These bioavailable contaminants may enter the detrital compartment of aquatic food webs, following plant senescence, or may be taken up directly by herbivores. Contaminants associated with plant debris drifting from Lake St. Clair and the Detroit River contribute a significant burden of bioavailable organic contaminants to the western basin of Lake Erie.

In contrast to their ecological importance, submersed rooted macrophytes have been overlooked in environmental science. Presently, the array of standard phytotoxic bioassays includes only one free-floating vascular macrophyte (Lemna) and several algal species. A short-term and inexpensive assay was studied for feasibility in evaluating sediment quality. Cloned ramets of the macrophyte Vallisneria americana were used to test phytotoxicity of sediments collected at different locations in the Detroit River. Ramets were planted in sediment samples and placed in greenhouse aquaria. After a week of exposure, ramets of V. americana were destructively sampled and preserved. The leaf and root surfaces areas were determined, and plant biomass was recorded for each ramet. An index of the leaf-to-root surface area ratio was a reliable predictor of sediment phytotoxicity; the ratio of leaf-to-root mass was also useful but proved less consistent. Ramets grown in sediments that were relatively less contaminated with organic compounds had lower values of the leaf-to-root surface area ratio, while plants grown in more contaminated sediments had greater values. Results of analyses of variance indicated that the index of leaf-to-root surface area ratio responded to sediment quality but was not significantly affected by either variation in plant genotype or interaction between sediment and plant genotype. There was a significant correlation (p < 0.001) between rank-ordered results of the present greenhouse study and results of leaf-to-root surface area ratios for plants previously surveyed in the field.

This study was carried out to assess the effects of sediment type, local water column, source of plants, and duration of exposure between 1990 and 1994 upon growth, survivorship and reproduction of Vallisneria americana at two sites in the Huron-Erie corridor of the Great Lakes. The major objective was to determine whether year-to-year variations affect consistency of results using Vallisneria as a biomonitoring tool. Relative annual rankings of results from experimental treatments did not change over a 4-year period. However, most of the absolute measures of plant performance varied significantly between years. Effects on plant growth and development were associated primarily with exposure to water columns, and secondarily to sediments. Measures of plant density, rate of clonal growth, leaf length, number of leaves per m 2 , and plant biomass per m 2 were unusually high in the first year compared with the subsequent 3 years, perhaps due to disturbance associated with set-up of the experiment. Leaf-to-root surface area ratios provided a simple, consistent, and reliable measure of environmental quality. Other measures of plant performance varied significantly from year-to-year and, therefore, are not advised for long-term comparisons. Results of this study indicate that Vallisneria can be an effective temporal biomonitor of environmental quality.

Clonal growth and reproduction in tristylous Lythrum salicaria L. were examined experimentally, using cloned genotypes of each of the three flower morphs, in field studies involving four moisture and three nutrient treatments. Clonal growth was measured in terms of diameter of clones, number of ramets per clone, and total length of ramets, and an index of reproduction was recorded as the total length of infructescence per clone. Neither clonal growth nor reproduction differed significantly among flower morphs, but both differed significantly as a consequence of both moisture and nutrient treatments. The pattern of seasonal growth indicates that ramet production was restricted mainly to the beginning of the season following vigorous vegetative growth. Although flowering began in June, it was restricted to plants in drier treatments in the water-gradient experiment. Characters intrinsic to tristyly (such as lengths of styles and stamens, and allocation of biomass to stamens and pistil) differed significantly among morphs. Soil moisture levels but not fertilizer treatments significantly affected the size of floral structures and biomass. Although absolute levels of biomass allocation to whole flowers and to attractive structures did not differ significantly among morphs, relative allocation to stamens increased progressively from long morph to mid-morph to short morph, with a corresponding decrease in relative mass of pistil. Although proportional allocation differed significantly among morphs, it was unaffected by moisture treatment, suggesting tight genetic control of herkogamy (spatial separation between anther and stigma). This should maintain the floral polymorphism in different ecological conditions. Key words: Lythrum salicaria, nutrient and water gradients, heterostyly, floral morphometry, floral allocation, clonal growth, sexual reproduction.

Aquatic macrophytes are useful biomonitors of sub-lethal effects of organochlorine contamination. Ramets of Vallisneria americana were sampled from natural populations at 243 sites in the Huron-Erie corridor in August 1993, and an index of leaf-to-root surface area ratios was determined. Sites were then ranked according to this index. Site rankings based on leaf-to-root surface area ratios were significantly correlated with the ranks of these sites using levels of organochlorine contamination of biota or sediment, reported independently in the published literature. The ratio of leaf-to-root surface areas ranged from 2 to 92, and higher values were found in more polluted sites. At four sites, plants were sampled monthly over one growing season, while at two sites plants were sampled each year for 4 years. The leaf-to-root surface area ratio changed little from year to year. At highly polluted sites there was a progressive increase in ratio over the growing season. Surveyed sites remained in the same relative ranking over the growing season, and a single survey within the same month could compare sites reliably. These results demonstrate that a very simple measure of ratios of leaf-to-root surface areas in Vallisneria may provide a rapid, sensitive, convenient and inexpensive metric of site quality from the perspective of macrophytes and organochlorine contamination. This could be used to prioritize sites for remedial action, and could confirm environmental improvement following remediation.

The value of wetland is now widely recognized; some legislation requires "no net loss" of wetlands, although economic incentives still exist for wetland conversion. Rare plants may be protected by law; however, wetlands are rarely managed specifically to conserve rare species. Furthermore, it is not always clear how the environment should be manipulated to increase the abundance of such species, since necessary autecological details are rarely available. Species conservation involves demographic and genetic elements, as well as ethical decisions about the merits of transplanting or importing genes through controlled pollinations. Rare species may serve as indicators of habitat quality, although this will depend on the reasons behind the species' rarity. There is a need for multiple-use management plans that incorporate species- and habitat-conservation goals and that implement overall strategies to maintain or enhance the total quantity and quality of wetlands. Key words: rare plants, management, wetland conservation, habitat conservation.

Gender phenotypes of Thymelaea hirsuta (L.) Endl. were recorded at five natural populations during regular visits occurring 3–6 times per year over a period of 3–6 years. Plant gender was quantifie...

Fruits of Vallisneria americana were collected in late September from four sites along the Huron-Erie Corridor, on the Canadian side of this connecting channel between Michigan and Ontario. Some fr...

In a number of disciplines including ecology, ecotoxicology, water quality management, water resource management, fishery biology etc., there is significant interest in the testing of new materials, environmental samples (of water or sediments) and specific sites, in terms of their effects on biota. In the first instance, we consider various sources of aquatic pollution, sources typically associated with developed areas of the world. Historically, much water quality assessment has been performed by researchers with a background in chemistry or engineering, thus chemical analysis was a dominant form of assessment. However, chemical analyses, particularly of such materials as organochlorines and polyaromatic hydrocarbons can be expensive, and local environmental factors may cause the actual exposure of an organism to be little correlated with chemical concentrations in the surrounding water or sediments. To a large extent toxicity testing has proceeded independently of environmental quality assessment in situ, and the work has been done by different, and differently-trained researchers. Here we attempt to bring together the various forms of biological assessment of aquatic pollution, because in our opinion it is worth developing a coherent framework for the application of this powerful tool. Biotic assessment in its most primitive form involves the simple tracking of mortality in exposed organisms. However, in most natural environments it is extended, chronic exposure to contaminants that has the most wide-ranging and irreversible repercussions--thus measures of sub-lethal impairment are favoured. From an ecological standpoint, it is most valuable to assess ecological effects by direct study of in situ contaminant body burdens and impairment of growth and reproduction compared with 'clean' sites. A distinction is made here between bioindication and biomonitoring, and a case is made for including aquatic macrophytes (angiosperms) in studies of contaminant levels and effects in the biota. It is apparent that there is a concurrent need for laboratory-based testing of new industrial by-products before any are released in the environment, and such studies should aid the investigation of mechanisms and modes of toxicity, but environmental assessment, and tracking of improvements in environmental quality are most effectively achieved by active biomonitoring experiments.

This paper assesses the value of American wildcelery, Vallisneria americana, as a biomonitor of organochlorine contamination. Plants provide a valuable indicator of local environmental conditions and sub-lethal effects of contaminants on growth and reproduction provide a means of assessing both acute and chronic contaminant effects. In a field survey, Vallisneria plants in the St. Clair and Detroit rivers were found to accumulate significant amounts of organochlorine contaminants, and the concentration increased over the season. Root tissue contained the highest concentrations on each occasion, suggesting contaminant transfer occurred from sediments to the roots. A factorial experiment was set up at two stations in the channel connecting Lakes Huron and Erie to assess the separate effects of contaminant concentrations in the sediment, water column, and source population of the plants, upon growth and reproduction of Vallisneria plants. Contaminant concentrations in sediment and plant tissues were also measured to correlate contaminant content and demographic effects. A clear relationship between exposure to contaminants and effect (on plant performance) was observed. Results indicated that exposure first to the water column, and secondarily to the sediments from the more contaminated site had detrimental effects on plant performance and survival. Plants at each site appeared to be locally adapted and differed in their apparent resistance to organochlorine contaminants. An “impairment index” (reflecting relative plant performance) was calculated and can be used to calibrate the degree of contamination of different sites relative to a “clean” reference site. This may allow prioritization of remediation of contaminated sites, and should allow field managers to track and document the restoration of environmental quality in recognized Areas of Concern in the Great Lakes.

Seed banks and sexual reproduction are known to be significant in colonization and re-establishment of some aquatic macrophyte communities. For highly clonal aquatic macrophytes, however, there is a lack of information on seed production and seed fate as compared with annual sexual species. The seed bank for three populations of Vallisneria americana in the Huron-Erie corridor of the Great Lakes was sampled and quantified in the spring of 1994, and related to seed production in the previous season at these sites. Seed deposition rates during 1994 were also assessed. Sites varied in the proportion of plants flowering and in their tertiary sex ratios, but did not differ in seed numbers produced per unit area. The size of the seed bank was not significantly related to the previous season's seed output, and estimates of seed deposition in the following year tended to be approximately tenfold greater than seed densities found in the seed bank. The stages between seed production and subsequent seed germination are generally very dynamic, with dispersal, mortality, and predation as likely regulating factors. The potential for seedling establishment in V. americana needs to be assessed more fully before the role of seeds in population processes can be determined.

Lythrum salicaria is a serious weed of wetlands in Canada, particularly in Ontario and Quebec. Indeed some assessments suggest that this exotic is now responsible for the conversion of more wetland habitat than is current human development pressure¡ Lythrum is a heterostylous species and clones by means of root buds. Despite its tristylous breeding system, Lythrum is a successful colonizer, and can produce prodigious numbers of seeds. It is likely that control can be achieved by hand-pulling in low-density populations, but once the species is established it generally becomes abundant and often approaches a monoculture. This species is particularly difficult to control because of its strong competitive ability, the extreme sensitivity of wetland habitats, and the likelihood of serious repercussions for wetland wildlife and fisheries if herbicides are used to control it. We see better possibilities for control through cultural techniques (e.g., manipulation of water levels to favour native species), enhancement of North American herbivores, and, possibly, the introduction of biological control agents from Europe and Asia.Key words: Purple loosestrife, Lythrum salicaria, wetlands, introduced weed

Ranunculus repens L. (creeping buttercup) is a common weed in the Maritimes, Ontario and Quebec, as well as British Columbia. A review of the literature on the biology of the species is presented. Ranunculus repens can develop large spreading clones as it produces ramets on stolons, and readily regenerates from root fragments. It is therefore particularly resistant to cultivation as a weed control strategy. It is, in addition, partially, or fully resistant to many herbicides, and its seeds show dormancy and sustained viability in the seed bank. Several investigations have been carried out on the population dynamics of this species, and its responses to light and nutrient treatments have been described.Key words: Weed biology, Ranunculus repens, creeping buttercup

In this review we focus on use of the aquatic macrophyte, Vallisneria americana, as a biomonitor of overall environmental conditions in the Laurentian Great Lakes. An array of measures of plant performance have been investigated; estimates of the leaf-to-root surface area ratio have proved to be the most consistently effective and useful. The species has been used in many different ways to characterize plant response to single organochlorines and metals, PCB mixtures, and as a bioassay of sediment toxicity, in the lab and in the field, to evaluate designated Areas of Concern, and to focus upon individual microsites and point source impact zones.

This study reports the bioconcentration and the uptake and elimination kinetics of a series of nonreactive, hydrophobic organic substances in the submerged aquatic macrophyte Myriophyllum spicatum. The tested substances represent a wide range of aqueous solubilities and 1-octanol-water partition coefficients (K{sub OW}). The plant-water bioconcentration factor is shown to follow a linear relationship with the octanol-water partition coefficient for all chemicals, including the superhydrophobic chemicals with log K{sub OW} up to 8.3. The uptake and elimination rate constants tend to follow a biphasic relationship with K{sub OW}. A kinetic model is developed for organic chemical bioconcentration is submerged aquatic macrophyte species. This model is applied to the Detroit River and Lake St. Clair to illustrate the role of aquatic macrophytes in chemical dynamics in aquatic systems.

We conducted a four-year field competition experiment involving Lythrum salicaria and the North American wetland species that it occasionally displaces, Typha angustifolia. The design placed Lythrum and Typha in a modified replacement series experiment having four starting densities (64, 36, 16, and 1 plant(s) per quadrat) and four relative proportions of each species (i.e., 0.25, 0.50, 0.75, or 1, [for densities 64, 36 and 16 only]). Results of a repeated measures ANOVA indicate that overall rate of ramet production (ORP) differed significantly between species, between density treatments and between years. We also found evidence of significant intra-specific competition in both species. In the first year of the experiment, ORP in Typha was greater than that in Lythrum. However, from the second year onward, this situation was reversed and ORP in Typha was much lower than that in Lythrum. Analysis of covariance indicates a significant effect of species proportion on ORP, suggesting that both intra-, and inter-specific competition were occurring. Log-transformed ratios of input/output (i.e. proportions of Lythrum/Typha established at the beginning of the experiment, divided by the proportions found at the end of each growing season) demonstrated that in the first year Typha gained advantage in the mixtures, while by the second and third years the output ratio line crossed the 45° threshold, suggesting the two species could coexist. However, by the fourth year, the output ratio line was entirely above the 45° line, suggesting an overall advantage to Lythrum in the mixtures. The experiment clearly demonstrates temporal development in competitive behaviour, and suggests that Lythrum may take at least four years before becoming competitively dominant in a wetland community. The results also suggest that short-term experiments should be interpreted with care, if they are to be used to forecast the long-term outcome of competition in nature.

Summary 1. The demographic effects of an organochlorine contaminant on the dioecious aquatic macrophyte Vallisneria americana are reported. Plants were sampled from two natural populations in the Huron-Erie corridor of the Great Lakes Basin, and placed in experimental treatments of high (396 mg per litre), medium (132 mg per litre), low (66 mg per litre) and zero concentrations of trichloroethylene (TCE). Plant responses were determined in terms of genet mortality (i.e. death of the entire genetic individual), ramet production (clonal growth) and mortality, leaf birth- and death rates, leaf area and sexual reproduction. 2. TCE in the water caused significant ramet and genet mortality, and reduced the growth of surviving plants. Control plants produced significantly more leaves than contaminated plants. Contaminant-exposed plants continued producing their leaves for 4-6 weeks longer than controls, and showed greater rates of leaf death. The total leaf area of genets exposed to TCE was significantly lower than that of unexposed plants. TCE was associated with a significant decrease in flower production. 3. The concentration of TCE was determined in plant tissues, water and sediment. Levels of TCE were greatest in the sediment, and sediment originating from each site adsorbed TCE to differing degrees. TCE was more concentrated in the underground storage tissues (roots and turions) than it was in leaves. Root and turion tissues of plants from one site contained significantly less TCE than those of plants from the other site, despite an identical concentration of TCE in the water; differences may have been due to contrasting sediment affinities for TCE, and/or different rates of TCE accumulation by the plants from each site. 4. TCE was bioaccumulated, particularly in roots. This appeared to be attributable to the fact that TCE accumulated, in the first instance, in sediments. The logIo of TCE concentrations in sediment were linearly correlated with TCE concentrations in roots, suggesting that TCE uptake is through the root system rather than the leaves, and that uptake is driven by the concentration gradient between sediment and plant. 5. Plants that survived the TCE treatment appeared to be resistant to the chemical; they were as likely to flower as surviving untreated plants. Males were more likely than females to survive TCE treatment. 6. It is concluded that modular demography can be used to characterize the performance of plants that are stressed by exposure to various concentrations of the organochlorine solvent trichloroethylene, and that plants from different sites differ in their ability to survive contaminant stress.

CONCEPTUAL ISSUES IN PLANT BREEDING SYSTEMS: J. and L. Lovett-Doust: Sociology of plants - an emerging synthesis Robert Bertin: Paternity in Plants David Haig and Mark Westoby: Inclusive fitness, seed resources, and maternal care Paul Cox: Monomorphic and dimorphic sexual strategies - a modular approach Spencer Barrett: The evolution, maintenance, and loss of self-incompatability systems Mark Schlessman: Gender diphasy ("sex choice") ECOLOGICAL FORCES: Michael Zimmerman: Nectar production, flowering phenology, and strategies for pollination Tom Lee: Patterns of fruit and seed production Don Waller: Plant morphology and reproduction Jacob Weiner: The influence of competition on plant reproduction Steve Hendrix: Herbivory and its impact on plant reproduction REPRODUCTIVE STRATEGIES OF NON-ANGIOSPERMS: Robert De Wreede and Terry Klinger: Reproductive strategies in algae Brent Mishler: Reproductive ecology of bryophytes Michael Cousens: Reproductive strategies of pteridophytes.

Individual higher plants may be male or female to a varying degree. The division of sexual labor leaves the maternal parent in any successful mating bearing the additional cost of nurturing embryos within developing seeds. Patterns of parental sex allocation reflect the evolutionary fitness that may be gained by assuming various gender roles. The evolution of gender separation is discussed. (Accepted for publication 15 August 1982)

Summary Experiments were performed to study sources of variation between ramets of a clone, and differences in growth form between plants from two locally specialized populations of Ranunculus repens in a woodland and a grassland site. In one experiment, ramets were taken from different positions along each of five primary branches of clones. For both populations variability in growth parameters of ramets of each clone was attributed to the position of a ramet within a branch, and not to any differences between branches of a clone. In clones which are made up of several interconnected ramets, these ramets may: (1) compete with each other for light and other resources; and (2) supply each other with resources. In clones where connections have been lost only (1) above, will pertain. In a second experiment clonal material selected from the two populations was grown in flats under conditions of high ramet densities. For each clone, in half of the replicates, stolon connections between ramets were severed; in the other half stolons were left intact. In clones originating from the grassland population, ramets of the same genotype accumulated significantly more total dry wt and had more above-ground biomass when they were interconnected than if stolons had been cut. In clones from the woodland population, total dry wt was not significantly altered. In plants from both populations, after connections were cut, the distribution of dry matter to various plant organs was significantly altered (plants distributed more dry matter to inflorescences and ramet caudices, and less to stolons). The pattern of proportionate allocation of dry matter suggested that below-ground competition was more intense when stolons had been severed. Growth form and the response to intraclonal competition are discussed in terms of local specialization in the woodland and grassland sites.

We describe the effect of repeated fruit harvest on yield and components of yield in bush bean, and we construct several growth models. Multiple-regression analysis was used to estimate the relative contributions of individual components to overall yield, and a technique comparing 'backward' and 'forward' multiple regressions was used to identify points of direct and indirect influence on the ontogenetic series contributing to yield. By testing alternative models we were able to identify components which did not contribute significantly to yield. Plant height was an important direct and indirect determinant of yield in both treated and untreated plants. There were more compensation effects between components in plants which were continuously harvested, perhaps because pods no longer contributed photosynthate, or because harvested plants were 'rejuvenated' through bud release, thus more resource sinks were competing actively for limited resources. Competition between resource sinks apparently occurs earlier in the model sequence where pods are harvested continuously, and becomes important later, as pods mature, in plants left intact to the end of the experiment. The reciprocal of leaf area ratio was another major component of yield. In both treated plants and controls, this ratio of leaf area to total plant dry weight behaved indirectly, apparently influencing yield through other components. THE YIELD OF A PLANT may be regarded as a consequence of sequential events. These events may correspond to developmental stages or to particular physiologic processes. Each event is seen as a function of preceding events, modified by environmental influences. These events, in total, are the components of yield. Sequential yield component analysis (SYCA) has identified significant components of yield in several species (e.g., Eaton and Kyte, 1978; Huxley, Brink and Eaton, 1979). Any component of yield may affect yield directly and/or it may act indirectly by influencing components which will be produced later in the developmental sequence. In SYCA, successive events are often highly correlated with each other. Because of their sequential nature, it is possible to consider the additional contribution to yield made by each event, after the value of the preceding one has been determined. SYCA through 'forward' analysis provides an estimate of direct and indirect action of components. The forward analysis I Received for publication 10 September 1982; revision accepted 7 December 1982. We acknowledge the financial support of Natural Sciences and Engineering Research Council of Canada Strategic Grant GO-407. We thank Pat Bowen and Barbara Neilson for helpful comments and technical assistance. gives a measure of the additional effect of any component, without separating the influence of ontogenetically earlier components. Bowen and Eaton (1983) have developed a method of 'backward' SYCA which in combination with the usual forward SYCA (e.g., Eaton and Kyte, 1978), can give an estimate of the direct effect of single components on yield. For example, any component is apparently not affected by earlier components if its coefficient of determination (R2) is O in a forward SYCA in which that component is itself taken as a dependent variable. If this component has a coefficient of determination greater than 0, when taken as an independent variable in 'backward' SYCA, then the effect of this component on yield is assumed to be direct (since its effects through chronologically later variables already have been accounted for and removed), and it is independent of chronologically earlier variables. The terms, backward and forward, as used here imply directions of stagewise inclusion processes (decided by the investigators), but it is worth pointing out that they have nothing to do with backward or forward stepwise regression. In stepwise regression these terms would reflect the use of either exclusion or inclusion algorithms, respectively, but where the sequence of exclusion or inclusion of variables

Arisaema triphyllum is a gender-labile woodland herb in which sex expression is correlated with the abundance of stored resources. Larger plants are female or monoecious, smaller ones are male. Among females larger plants produce more flowers, fruits and seeds, and the rate of successful fruit and seed formation is greater for plants of greater ht and corm diam. Average seed wt is greater in larger plants. Seed number per fruit and average seed wt per fruit taper towards the top of the infructescence. Pollinator limitation and resource supply may both contribute to the regulation of yield; their effects can be interpreted sequentially. SEX EXPRESSION IN Arisaema triphyllum is labile and is correlated with plant size and the abundance of stored reserves in the underground corm (Treiber, 1980; Policansky, 1981; Lovett Doust and Cavers, 1982; Bierzychudek, 1 984a). Larger plants are female (or monoecious) while smaller plants are male or nonflowering. Components of plant fitness such as flower number, and rates of fruit and seed set are closely linked to plant size (for example ht or leaf area). Policansky (1981) pointed out a significant correlation between female fecundity and plant size in A. triphyllum but found no such correlation for males. It has been suggested that a paucity of insect pollinators sets a low limit to the reproductive success of Arisaema triphyllum (Treiber, 1980; Bierzychudek, 1981). Bierzychudek had observed that seed production in A. triphyllum at two populations was not a function of plant size, prompting her pollinator limitation hypothesis (198 1). If we compare the results of different workers, it is apparent that there is considerable variation in seed set from place to place and from year to year (Table 4). Fecundity in A. triphyllum, like sex expression in this species, may also be regulated by the quantity of stored resources. The resources remaining in the corm in fall should be a good predictor of fecundity. Other measures of plant size, such as ht or above-ground biomass, may ' Received for publication 30 April 1985; revision accepted 30 October 1985. We thank Mount Holyoke College for their 1983 faculty research grant to LLD. Jane Leopold carried out preliminary work that was valuable in formulating some of the fruit analyses reported here. David Policansky, Paulette Bierzychudek and Spencer Barrett provided very helpful reviews of an earlier draft. 2 Present address: Amherst College, Department of Biology, Amherst, MA 01002. also correlate closely with yield. In this paper we are concerned with aspects of seed set in jack-in-the-pulpit. We estimate components of fecundity and relate these to plant ht and corm diam, as well as fruit position within the infructescence. MATERIALS AND METHODS-In September 1982, 24 femalejack-in-the-pulpit plants bearing fruits were sampled from the field. Plants were selected randomly from among all females that were fruiting. (Hence, this represents a significant difference from the sampling methodology of Bierzychudek [e.g., 198 1, 1 984a, b] who sampled randomly from among all plants that flowered as females, and would therefore have included some females that set no fruits at all.) The infructescence was removed with the rest of the stem, and the corm and roots were excavated. Plant ht was measured from the corm to the highest part of the plant, and corm diam was measured at its widest part. Individual fruits were removed from each infructescence; seeds were removed from each fruit and their number recorded. Air-dried seeds were weighed. Seeds were tested for germinability and those that did not germinate were tested for viability using tetrazolium chloride. A further 25 infructescences were collected for analysis the following year. The ht of the infructescence and the diam of the peduncle at the base of the infructescence were measured. Individual fruits were categorized as being from the bottom third of the inflorescence, the middle portion, or the top third. Fruits were counted and weighed, and the number and wt ofindividual seeds within fruits from these plants were determined. RESULTSIntact plantsMeans and S.E. for a variety of aspects of growth and yield are

The hypothesis that male-biased sex ratios result from the greater costs of reproduction for females was examined at various modular levels in a study of the dioecious clonal shrub, staghorn sumac (Rhus typhina). The population of flowering trunks was male- biased, though male and female clones proved to be of similar ages. Trunks within apparent female clones were more likely to be vegetative or dead than in male clones, though the density of living plus dead trunks was comparable in male and female clones. Females of given ages had the same mean trunk diameter as similarly aged males. Overall, the average female trunk diameter was the same as the average male diameter. Differential costs and patterns of reproduction for males and females were revealed through destructive harvests and separate analysis of trunks and branches. An abrupt discontinuity in the performance (growth) of male trunks was found; those with a diameter less than 4 cm had greater total numbers of leaves, branches, and inflorescences than females of comparable diameter, whereas larger males and females did not differ from each other. Flowering male branches had significantly fewer leaves than those of flowering females. The number of leaves per branch was highly correlated with the number of fruits produced in the corresponding terminal inflorescence. The average (? SE) male inflorescence had 4099 ? 312 flowers, whereas the average female inflorescence bore 1575 ? 67.2 fruits. There was a sixfold difference in the biomass of these terminal reproductive units (female infructescences being heavier). This study suggests the importance in a clonal species of defining the modular level at which costs are assessed. The hypothesis that females suffer greater reproductive costs as a consequence of fruit production is supported by the finding of diminished survivorship of daughter trunks in female clones. However, it is not supported by data on annual trunk diameter increments. Reproductive costs in sumac are shown at the level of the clone (ramet survivorship) or the branch, rather than at the level of the flowering trunk module.

Reciprocal pollinations with control treatments were made within and among four closely related taxa of Oxalis: O. corniculata, O. stricta, O. dillenii ssp. dillenii, and O. dillenii ssp. filipes.The following conclusions were made. (1) There was evidence that O. d. ssp. filipes was capable of seed set by facultative apomixis. (2) All four taxa were self-compatible and showed highest seed set when left to self-pollinate. (3) In all matings seed set occurred, but the level of fecundity was dependent upon species and also upon whether a species served as maternal or paternal parent. (4) The seeds were subsequently germinated. It was found that measures of intertaxon compatibility based on capsule formation or seed set alone did not necessarily reflect the level of interfertility in terms of seed viability. (5) Differential reproductive success of each mating generally supports the earlier taxonomic treatment of Eiten for the section.

British Umbelliferae form a natural cohesive taxon, having patterns of habitat distribution, reproductive biology and life-history that distinguish them from other British dicotyledons. They are usually found in open sites, and are local or rare in distribution. Almost half of British species have the andromonoecious sex habit. Perfect flowers are usually protandrous. The floral composition of umbels, and the pattern of opening of the inflorescence combine to produce a protogynous effect. These have been interpreted as mechanisms which decrease competition for resources between male and female functions in the plant, such that separate maternal and paternal fitness can become optimal. Seeds often require a chilling treatment, alternating temperatures, or a period of after-ripening before they are capable of germination. Schizocarps are of low weight. This seems to relate to the relative openness of habitat wherein unbellifers are found. Patterns of dry matter distribution reflect the life-history strategies of various umbellifer species. Agromyzid leaf miners parasitize the more widespread species, and those which are found in a greater number of different habitats.

SUMMARY (1) Demographic investigations were made over a period of 18 months in two adjacent populations of the clonal perennial Ranunculus repens in park grassland and in mixed deciduous woodland in North Wales. Ramets were treated as units of population. (2) The carrying capacity of the woodland site for Ranunculus repens (inferred from the peak summer density of ramets) was twice as high as that of the grassland site (264 and 112 ramets m-2, respectively, in 1977). (3) Despite the presence of a large and viable seed bank in the grassland soil (>1000 seeds m-2), germination and establishment of new genets was rare in both sites. (4) The birth rate of ramets per rosette was apparently density-independent, but death rates per rosette were density-dependent, particularly in summer. (5) Clonal growth and the death of ramets were in phase throughout the study. The average production of new ramets by each rosette was similar at both sites (4.6 daughter ramets in woodlands, 4.0 daughter ramets in grassland). (6) Demographic treatment assumes that the units enumerated may validly be compared. This assumption was tested and daughter ramets were found to have the same biomass at both sites, but daughter ramets were differently constructed at the two sites; those in grassland had proportionately more dry matter in caudex and root tissue than those in the woodland, which had more biomass in stolons. (7) The terms 'phalanx' and 'guerilla' are introduced to describe patterns of clonal growth, and the two populations of Ranunculus repens are evaluated in these terms.