• Tracey, Amanda J.
  • Aarssen, Lonnie W.



According to traditional theory, superior competitive ability in plants generally requires relatively large plant body size. Yet even within the most crowded vegetation, most resident species are relatively small; species size distributions are right-skewed at virtually every scale. We examine a potential explanation for this paradox: small species coexist with and outnumber large species because they have greater ‘reproductive economy’, i.e. they are better equipped—and hence more likely—to produce offspring despite severe size suppression from intense competition.


Randomly placed plots within old-field vegetation were surveyed across the growing season. Within each plot, the largest (MAX) and smallest (MIN) reproductive individuals of each resident species were collected for above-ground dry mass measurement. We tested three hypotheses: (i) smaller resident species (with smaller MAX size) have generally smaller reproductive threshold sizes; (ii) smaller resident species have greater ‘reproductive economy’, i.e. a smaller MIN relative to MAX reproductive plant size; and (iii) MIN size predicts plot occupancy (species abundance within the community) better than MAX size.

Important Findings

The results supported the first and third, but not the second hypothesis. However, we could not reject the hypothesis that smaller species have greater reproductive economy, as it was not possible to record data for the largest potential plant size for each species—since even the largest (MAX) plants collected from our sampled plots were subjected to competition from neighbours under these natural field conditions. Importantly, contrary to conventional competition theory, more successful species (in terms of greater plot occupancy) had smaller minimum not larger (or smaller) maximum reproductive sizes. These results suggest that a small reproductive threshold size, commonly associated with relatively small potential body size, is generally more effective in transmitting genes into future generations when selection from neighbourhood crowding/competition is intense—at least within natural old-field vegetation. Accordingly, we propose a simple conceptual model that represents the basis for a fundamental paradigm shift in the predicted selection effects of crowding/competition on plant body size evolution.


Study site

The field work was conducted from June to August 2009 at the Queen's University Biological Station, near Chaffey's Locks, Ontario, Canada (44°33′N, 76°21′W), in a small old-field community known locally as the ‘Wire Fence’ field. The field is irregular in shape, ∼2 ha in size and is surrounded on all sides by mature woodland. The field was last tilled and sown about 70 years earlier with a hay mixture of Phleum pratense L. and Trifolium pratense L. and has been mown periodically for hay since that time. Aside from this, the field has been largely undisturbed by humans except for occasional foot traffic from hikers and field workers, and vegetation surveys and specimen collection by researchers at the field station. Resident species included mostly perennial dicots, grasses and sedges typical of old-field vegetation in the south-eastern Ontario region. The most common species in the field (measured in terms of frequency of occurrence in plots) were P. pratense L., Poa pratensis L., Vicia cracca L., T. pratense L. and Ranunculus acris L.

Data collection

In order to include data for both earlier- and later-flowering species, four vegetation surveys were conducted—spaced out across the growing season: from June 1 to 5; from June 22 to 29; from July 27 to August 5 and from August 20 to 27. In each survey, 1 × 1 m plots were positioned randomly within the field, and sample sizes included as many plots as could be processed over a 5-day period. A total of 155 plots were sampled throughout the entire growing season. Sampled plots included only those with a minimum of two resident species and a minimum of two reproductive individuals per species. A reproductive individual was defined as one displaying flowers or indications of present or recent seed/fruit production. Within each plot, all resident species were identified and recorded, and the largest (MAX) and smallest (MIN) reproductive plant of each species were clipped at ground level and collected in separate paper bags (below ground biomass was not collected because it was virtually impossible to accurately separate root material). In order to be certain that the smallest reproductive individual was obtained, it was necessary to harvest virtually the entire plot, carefully inspecting each plant. For tufted, non-creeping sedges and grasses (e.g. P. pratense L.), the whole tuft of tillers was collected as an individual plant. For creeping (rhizomatous) species (e.g. P. pratensis L.), just the local tiller or tuft of tillers visible above ground—a ‘rooted unit’ (Aarssen 2008)—was collected and regarded as one ‘individual’. For stoloniferous species (Fragaria virginiana L.Trifolium repens L.), a rooted unit that could be seen attached above ground was collected and regarded as one individual plant. Harvested plants were dried to a constant final mass and were weighed individually to obtain total above-ground dry mass per plant.