摘要:

Island biogeography theory attempts to explain and predict among—island variation in species richness. However, two islands with the same number of species may still differ from each other considerably in their species composition. In this study we test the hypothesis that among—island variation in species composition is predictable and can be related to the corresponding differences in distance to the mainland. We focus on woody plants inhabiting islands in the Clarks Hill Lake, a reservoir completed in 1954 on the Savannah River, between Georgia and South Carolina, USA. Two groups of islands were sampled: islands that wee logged prior to the filling of the reservoir and islands that were not logged. Each island was surveyed for the presence of all tree and shrub species, and its distance from the mainland was determined. In both groups of islands, the degree to which two islands are similar in their species composition was negatively and significantly correlated with their difference in distance to the mainland. Species richness, however, was correlated with distance to the mainland only on logged islands. We conclude that geographic isolation may affect species composition on islands, and that such an effect may occur even in the absence of a corresponding effect on species richness.

ISSN:0012-9658    

DOI:10.2307/1940467

出版商:Ecological Society of America    

年代:1993

数据来源: WILEY

摘要:

The post—glacial history of two adjacent sites in the Harvard Forest, a 10—ha swamp (Black Gum Swamp) and a 0.006—ha hollow (Hemlock Hollow) in a Tsuga canadensis forest were investigated using pollen analysis. The sites were selected in order to contrast the regional vegetation history revealed from the swamp sediments with the local history of the Tsuga forest reconstructed from the Hollow sediments. Specific objective were (1) to document the natural and anthropogenic disturbance history, (2) to examine the long—term vegetation dynamics of the two sites resulting from environmental change, species migration, and disturbance, especially with respect to Tsuga, and (3) to contrast the pre— and post—settlement vegetation and environments. The Swamp and Hollow cores contain continuous sediment records covering the past 12 300 and 9500 yr, respectively. Regional vegetation changes are delimited in six pollen zones: I, Herb zone (12 500—11 800 yr BP); II, Picea zone (11 800—9500 yr BP); III, Pinus—Quercus zone (9350—8350 yr BP); IV, Tsuga—northern hardwoods zone (8350—1750 yr BP); V, Tsuga—Castanea—hardwoods zone (1750—200 yr BP); and, VI Post—settlement zone (200 yr BP—present). No disturbances are detected in the periods of tundra or boreal vegetation from 12 500 to 8 350 yr BP. Since 8350 yr BP three distinct disturbance processes are detectable: (1) fires recorded in discrete charcoal horizons, (2) the apparent pathogenic decline of Tsuga (4700—3500 yr BP) and the blight of Castanea (° 1915 A.D.), and (3) post—settlement forest cutting, burning, land clearance, and cultivation (1750 A.D. to present). Log—ratio diagrams facilitate the differentiation of local vegetation history in the Tsuga forest from the regional upland history. Tsuga has been the dominant taxon at the Hollow for the last 8000 yr although its abundance has fluctuated with disturbance. Tsuga, Ulmus, Populus, Acer rubrum, A. saccharum, Betula, and Castanea apparently are more abundant locally than in the regional vegetation, which has a larger component of Pinus, Quercus, and Carya. Post—disturbance vegetation dynamics in the Tsuga forest are controlled by the type and intensity of disturbance and the pool of available species, determined by species migration and climate change. Tsuga decreases and subsequently recovers in 300—1200 yr following most disturbances. Northern hardwood taxa (Acer saccharum, Fagus, Betula spp., Ulmus, Fraxinus) generally decrease after fire and human activity and recover with Tsuga. For the period 8000—3000 yr BP Pinus, Quercus, and occasionally Populus and Nyssa increase following fire and the Tsuga decline at 4700 yr BP. However, since its immigration at °3000 yr BP Castanea has been the major species to increase upon disturbance of the Tsuga—northern hardwood forest. The study documents the long—term dominance of a forest stand by Tsuga canadensis and northern hardwood species despite repeated infrequent disturbances. The ability of these taxa to gradually reassume their former abundance following disturbance is presumably a consequence of their shade tolerance and longevity and the low frequency of disturbance.

ISSN:0012-9658    

DOI:10.2307/1940468

出版商:Ecological Society of America    

年代:1993

数据来源: WILEY

ISSN:0012-9658    

DOI:10.2307/1940469

出版商:Ecological Society of America    

年代:1993

数据来源: WILEY

摘要:

Communities subject to disturbance are apparently vulnerable to invasion in that they contain a large fraction of the introduced plant species persisting in any given area. Still, ecologists have a poor understanding of how details of the disturbance regime influence the success with which invaders spread through a community. In this paper, we report results from an experiment that investigates how the spatial distribution of bare ground influences the rate at which offspring of an introduced invader spread through a perennial ryegrass community. Artificially created gaps of three sizes were positioned along linear transects according to one of two spatial distributions. These size transects (one for each treatment combination) were positioned as spokes on a wheel and Senecio vulgaris was introduced into the center. After controlling for the total amount of gap area, we monitored the number and position of Senecio vulgaris establishing for two subsequent generations. Our results showed that the rate of spread of Senecio vulgaris was sensitive to both the gap size and the gap distribution; plants moved a greater distance when the gaps were large and underdispersed. Simulation models confirm that this pattern follows from the way in which gaps sample from seed shadows. We derive an analytical expression to explain why the mean seed dispersal distance does not predict the spread of Senecio vulgaris adults, and show how rates of spatial spread will change under different spatial patterns of disturbance. Finally, a model of spatial point processes confirms that the distribution of gaps mediates the rate of weed spread in two dimensions as well.

ISSN:0012-9658    

DOI:10.2307/1940470

出版商:Ecological Society of America    

年代:1993

数据来源: WILEY

摘要:

The purpose of this study was to compare forest structure on a series of geologically similar larva flows that differed in age, but not in climate or in accessibility to pioneer plants. The data were then used to infer the patterns of forest development over 3000 yr on a single flow. Stand structure was measured for populations of Metrosideros polymorpha, other canopy tree species, and tree ferns on a chronosequence of lava flows aged 47, 137, °300, °400, and °3000 yr, in the montane rain forest zone on Mauna Loa, Hawaii. The upper stratum of vegetation at all sites was dominated by M. polymorpha; populations of other trees were relatively sparse throughout the study area. For M. polymorpha populations, the basal area per hectare increased, at an ever—decreasing rate, as flow age increased; population densities of seedlings, samplings, and trees peaked on the 137—yr flow, then declined, apparently through self—thinning and wind—throw, with further increases in flow age. The decrease in M. polymorpha population density on older flows was accompanied by an increase in the density of tree ferns, which formed a closed subcanopy on the 3000—yr flow, and may have inhibited regeneration of M. polymorpha. Among M. polymorpha trees there was a successional transition from pubescent varieties on the three youngest flows to glabrous varieties on the oldest flow. A comparison between plant populations on the chronosequence and on one other flow, aged °1400 yr but containing some volcanic ash, indicated that ash may alter some aspects of forest development.

ISSN:0012-9658    

DOI:10.2307/1940471

出版商:Ecological Society of America    

年代:1993

数据来源: WILEY

摘要:

Allometric theory on mechanisms of the self—thinning rule was tested for Nothofagus solandri populations from the Craigieburn Range, New Zealand and for Pinus densiflora stands from northern Japan. The self—thinning rule describes a consistent relationship of mean plant mass to the approximately —3/2 power of plant density in evenaged monocultures. Although his rule has been described for various species, mechanisms that produce certain relationships have not been well understood. We tested an allometric theory of Long and Smith of the self—thinning rule that assumes constant foliage mass density and allometry for mean dimensions of populations that represent dense conditions for given mean plant sizes. Only stands at maximum crowding were selected for analysis. The self—thinning boundary of N. solandri showed an exponent —1.13 with a 95% ci of —1.25 to —1.02 for mean stem mass. This was significantly shallower than the conventional value of the exponent —3/2, but was identical to the predicted exponent from the allometric theory. The thinning coefficient was also explained numerically by this hypothesis. In contrast, analysis of published data for P. densiflora indicated that the thinning exponent did not differ from the proposed —3/2. Empirical thinning lines varied substantially depending on species and plant parts considered; however, the allometric theory consistently provided predictions that agreed with the observed thinning relationships. Implications for the geometry of self—thinning populations and generality of the allometric theory are discussed.

ISSN:0012-9658    

DOI:10.2307/1940472

出版商:Ecological Society of America    

年代:1993

数据来源: WILEY

摘要:

Life history theory assumes that there is a trade—off between current reproduction and future growth or reproduction or both; therefore, natural selection is expected to result in the maximization of reproduction within the limits imposed by the trade—off. Accordingly, it has been predicted that fruit and seed production in perennial plants should be resource limited. Many orchid species seem to be pollinator limited, but this hypothesis has been recently challenged by experimental studies that show a cost of increased fruit set in some orchids. In this study, I combined the results of a 2—yr pollination experiment and a 3—yr demographic assessment of a population of the orchid Tolumnia variegata in a matrix model of population dynamics, and by means of simulations I evaluated the effect of pollination intensity and frequency, fruit production, and cost of fruiting on the asymptotic population growth rate. Mean natural fruit set in this population was<1%, whereas intermediate and high pollination intensity resulted in mean fruit set of 35 and 72%, respectively. In any given year, °98% of all flowering individuals fail to set fruit under natural conditions. Despite the dramatic increase in fruit set after hand—pollination, only plants in the high pollination intensity treatment showed a statistically significant reduction in future growth and flowering. The simulation showed that a small seedling production per fruit would be enough to overcome the cost of fruiting; therefore, plants in this population should experience strong selection for increased pollination. This conclusion does not seem to explain the widespread occurrence of low fruit set and high proportion of fruiting failure among nonautogamous orchids. The results of this study suggest an alternative explanation: pollinator limitation could be evolutionarily stable if the correlation between fruit or seed production and seedling recruitment is sufficiently low. If an increase in fruit or seed production does not translate into an increase in fitness, then selection for increased pollination would not occur or would be too weak. The evaluate this alternative, quantitative studies on the transition from seed to seedling in natural orchid populations are required.

ISSN:0012-9658    

DOI:10.2307/1940473

出版商:Ecological Society of America    

年代:1993

数据来源: WILEY

摘要:

Using experimental manipulations, we examined pollen and resource limitations to seed production and their interaction in a natural population of a monocarpic plant, Ipomopsis aggregata. Our design cross two factors, pollen level (hand—pollinated or control) and resource level (water added, water and fertilizer added, or control). Both hand—pollination and fertilizing during the blooming season increased total seed production, while watering alone has no effect on any component of reproductive success. Hand—pollination boosted number of seeds per flower, with no effect on flower number. In contrasts, fertilizing had its primary effect on the number of flowers produced, while also increasing the number of seeds per mature fruit in hand—pollinated plants. Fertilizing increased nectar volume, but path analysis detected no indirect effect of this increased reward rate on seed production, suggesting instead that fertilizing has a slight, but direct, effect on seeds per flower. These results argue against a strict dichotomy between pollen limitation and resource limitation of female reproductive success in plants.

ISSN:0012-9658    

DOI:10.2307/1940474

出版商:Ecological Society of America    

年代:1993

数据来源: WILEY

摘要:

We examined the role of growing season length in regulating percentage cover and seed mass of the alpine snow buttercup, Ranunculus adoneus, in a snowbed habitat. Season length varies from °50 d in early—melting portions of the snowbed to 35 d in late—melting locations. Percentage cover and seed mass of snow buttercups decrease from early— to late—melting locations. Time available for growth was manipulated experimentally by altering the snowmelt schedule. Advancing snowmelt by 14 d in typically late—melting portions of the snowbed led to a threefold increase in percentage cover of snow buttercups (P<.025). However, delaying snowmelt to the same extent in typically early melting locations had little influence on snow buttercups. These results indicate that time available for growth limits cover of snow buttercups in late—emerging vegetation, whereas other factors regulate cover of snow buttercups in early—emerging vegetation . Perturbation of snowmelt scheduled also had significant effects on seed mass (P<.0055). These effects are probably mediated through changes in time available for seed maturation. Snow manipulation treatment accounted for 71% of the variation in seed maturation interval. Maturation interval, in turn, was correlated with seed mass (P<.01). Reducing or increasing the seed maturation interval by 12 d caused a 33% change in seed mass. These findings show strong direct effects of growing season length on plant performance at diverse life history stages.

ISSN:0012-9658    

DOI:10.2307/1940475

出版商:Ecological Society of America    

年代:1993

数据来源: WILEY

摘要:

Stamens produce male gametophytes and expose them to the pollination process, so that stament characteristics should reflect selection on male function. For animal—pollinated plants, restricting removal in favor of using all available pollinators should generally promote pollen dispersal, so that staminal characteristics should limit pollen removal by individual pollinators. Here we describe the influences of anther position on pollen removal from Pontederia cordata, a tristylous plant with anthers at three distinct positions: exserted from the flower (long—level anthers), at the mouth of the flaring tubular perianth (mid—level anthers), and near the perianth based (short—level anthers). To measure pollen removal, we presented unvisited flowers to freely foraging bees and counted the number of pollen grains left in flowers after 1—4 visits. The pollinators included two generalist bumble bees (Bombus impatiens and B. vagans: Apidae) and Melissodes apicata (Anthophoridae), a specialist with morphological and behavioral adaptations for collecting pollen from P. cordata. Analysis of pollen removal incorporated the effects of bee species, anther position, pollen size, flower length, and the number and cumulative duration of bee visits. Anther position and bee species significantly influenced pollen removal, although the differences between bee species were not related to pollinator specialization. Long—level anthers dehisced early, lost an average of 39% of their pollen during their first visit, and did not consistently contact small pollinators, such as M. apicata. Short—level anthers restricted pollen removal more (28% lost) than long—level anthers, but dehisced more slowly and were subject to unpredictable pollen removal. Mid—level anthers combine the advantages of more exserted and inserted positions without suffering the corresponding disadvantages as they restricted removal (25% lost), dehisced relatively early, and seemed to consistently contact all pollinators. The overall benefits of positioning anthers within the perianth mouth, the position occupied by mid—level anthers in P. cordata, may explain the prevalence of this morphology among bee—pollinated plants with flaring, tubular, monomorphic flowers.

ISSN:0012-9658    

DOI:10.2307/1940476

出版商:Ecological Society of America    

年代:1993

数据来源: WILEY