Seed plants commonly allocate main stem growth to vertical (and in some cases horizontal) extension at the expense of allocation to growth of side branches. Referred to as apical dominance, this directional growth form ‘strategy’ is enabled by effects of the plant hormone auxin on suppression of the main stem’s ‘bud bank’ (Cline 1994). The bud bank is the individual’s population of axillary meristems positioned along plant shoots, and from which essential structures develop. In angiosperms, these meristems may have one of three fates (Fig. 1) (Bonser and Aarssen 2003): inactive (I) meristems remain in a dormant state and do not produce any structures; reproductive (R) meristems produce flowers or inflorescences and growth (G) meristems produce branches bearing leaves. G meristems (as well as the apex of the main stem) can terminate in one of two fates: either in a reproductive or an inactive apex. The former produces a flower or an inflorescence whereas the latter typically produces nothing. Whether the inactive or reproductive meristems are axillary or apical can be signified by the subscripts ‘ax’ and ‘ap’, respectively (i.e. Iax, Iap, Rax and Rap) (Bonser and Aarssen 2003) (Fig. 1). For a plant with strong apical dominance, many or most of the axillary buds found on the main stem remain in their inactive (Iax) form for most or all of the growing season.
Illustration of the three potential axillary meristem fates. In (a), an inactive (Iax) meristem (red circle) remains in a suppressed state (producing nothing), thus leaving resources available for potential allocation to directional main stem growth (e.g. height extension). In (b), an Iax has developed as a reproductive (Rax) meristem, producing a flower or inflorescence, thus promoting seed offspring production. In (c) an Iax has developed as a growth (G) meristem, producing a lateral leaf-bearing branch/shoot, thus promoting biomass accumulation and supporting three additional Iax meristems, and one terminal inactive shoot apical meristem (Iap) (red circles).
Apical dominance may have several fitness advantages for the plant, but these may also be associated with trade-off costs. The research reported here uses a multispecies field experiment to explore recent hypotheses concerning these potential benefits and costs of apical dominance. The benefit (adaptive advantage) of apical dominance may involve several interpretations (Aarssen 1995). The Light Competition hypothesis suggests that strong commitment to vertical growth (height) allows individuals in crowded vegetation to overtop close neighbours, thus also minimizing the probability of being shaded by them (Aarssen and Irwin 1991). The fitness advantage here assumes therefore that competition for light imposes a relatively strong selection pressure (which of course may not be the case in some habitat types).
The Effective Pollination hypothesis suggests that taller plants (promoted by apical dominance) may have an advantage in attracting pollinators (or in receiving/donating pollen via wind) relative to shorter neighbours, thus promoting greater reproductive success (Aarssen 1995; Donnelly et al. 1998)—including possibly with more outcrossing (less selfing) and hence less potential cost from inbreeding depression (Donnelly et al. 1998; Lortie and Aarssen 1999). As with the Light Competition hypothesis, the potential fitness benefits here may be more likely in habitat types where several resident species are competing for the same pollinators (although selection might also result from competition for pollinators between genotypes within a species).
The Effective Dispersal hypotheses suggest a potential benefit of apical dominance through access to effects of wind, which are generally stronger at higher distances above ground. Stronger winds will generally allow seeds to disperse farther away from the parent plant, even for species without specialized seed/fruit appendages that promote interception of wind currents (and especially for species that lack other seed dispersal mechanisms). This effect decreases the likelihood of competition between parent plants and their progeny, and also extends the probability of offspring colonization to nearby suitable habitat (Aarssen 1995). Wider dispersal also minimizes susceptibility of offspring to density-dependent seed and/or seedling predation (Howe and Smallwood 1982).
The final hypothesis—the Reserve Meristem hypothesis—proposes that the benefit of apical dominance lies in its effects on delaying release of axillary meristems from their inactive state (i.e. prolonging their Iax status), thus making them available (in reserve) for deployment should the plant experience apical herbivory (Aarssen 1995), and therefore enabling compensatory branch production for the afflicted plant (Lortie and Aarssen 2000a). The Reserve Meristem hypothesis may be more widely applicable (being independent of variation in competition intensity, or pollination/seed dispersal mechanisms). But importantly, its benefit is dependent upon being released from apical dominance, and so it assumes that the species is at least periodically (and predictably) susceptible to apical herbivory. In contrast, the other hypotheses for apical dominance involve benefits that are essentially passive and direct—i.e. incurred as a consequence of not being released from apical dominance.
There is no reason of course to generally expect these potential benefits of apical dominance to be realized individually. The mechanisms of two or more hypotheses may interact, or may simultaneously account for patterns of variation in the potential fitness consequences of apical dominance.
When the shoot apical meristem of plants is damaged or removed, fecundity and/or plant growth may suffer (under-compensation), remain unaffected (compensation) or increase (overcompensation). The latter signifies a potential ‘cost’ of apical dominance. Using natural populations of 19 herbaceous angiosperm species with a conspicuously vertical, apically dominant growth form, we removed (clipped) the shoot apical meristem for replicate plants early in the growing season to test for a potential cost of apical dominance. Clipped and unclipped (control) plants had their near neighbours removed, and were harvested after flowering production had finished but before seed dispersal. Dry mass was measured separately for aboveground body size (shoots), leaves, seeds and fruits; and number of leaves, fruits and seeds per plant were counted. We predicted that: (i) our study species (because of their strong apically dominant growth form) would respond to shoot apical meristem removal with greater branching intensity, and thus overcompensation in terms of fecundity and/or biomass; and (ii) overcompensation is particularly enabled for species that produce smaller but more leaves, and hence with a larger bud bank of axillary meristems available for deployment in branching and/or fruit production. Widely variable compensatory capacities were recorded, and with no significant between-species relationship with leaf size or leafing intensity—thus indicating no generalized potential cost of apical dominance. Overall, the results point to species-specific treatment effects on meristem allocation patterns, and suggest importance for effects involving local variation in resource availability, and between-species variation in phenology, life history traits and susceptibility to herbivory.
Study sites and species selection
The field work was conducted during the 2017, 2018 and 2019 growing seasons (May–October) in the vicinity of the Queen’s University Biological Station (QUBS), near Chaffey’s Locks, Ontario, Canada (44°13′ N, 76°36′ W), involving vegetation growing naturally along roadside and woodland edges, and within old fields and recently disturbed, abandoned habitats. Sampled populations had a minimum of 20 intact individuals, showing no, or negligible evidence of aboveground herbivory. Only species with a strongly apically dominant growth form were selected for study, but were otherwise chosen haphazardly as suitable populations were discovered.
Replicate plants (Table 1) within each population were randomly assigned to treatment and control groups. In order to account for effects of natural variation in plant size prior to treatment, each individual in the treatment group was randomly paired with an individual from the control group that was of a similar height (i.e. <4 cm difference; exceptions being Arctium lappa and Lactuca biennis which had maximum differences of 12 and 38 cm, respectively, due to limited availability).
Individuals in the treatment group had their main stem terminal apex (bud)—as well as its first (or first pair of) visible subtending leaf (leaves)—removed to simulate local (apical) herbivore damage. All neighbouring plants within a 25 cm radius around each individual (in both the treatment and control groups) were then cut at ground level in order to control for potential variation in shading effects from near neighbours. The initial height and number of leaves for each individual (in both the treatment and control groups) were also recorded before clipping (Table 2).
For about half of the species, the same treatment protocol was used for subsequent treatment events (Table 1), separated by about 3–4 weeks, using new individuals in the same populations. This was done in order to test for effects of variation in plant size at the time of clipping, and effects of variation in growing-season-time available for regrowth and reproduction.
Surviving, undamaged plants were harvested when flowering was finished and fruits were maturing and beginning to disperse seeds. Individuals were clipped at ground level, placed in plastic bags and stored in the laboratory at −4 °C for subsequent processing.
Each harvested plant was processed to record counts for each meristem and apex type (G, Iax, Rax, Rap and Iap) (Fig. 1). Additionally, counts were recorded for number of fruits, number of flowers and number of missing fruits (evident from peduncles). Following that, number of leaves adhering to the main stem, number of leaves adhering to any side branches and number of missing leaves (evident from leaf scars) were also recorded. All intact leaves adhering to the main stem, all intact leaves adhering to any side branches, the main stem, the side branches and all fruits were placed in separate paper bags, dried at 72 °C for 3 days, and then weighed to record dry mass values. Any harvested plant that was found to have been depredated prior to harvesting was not processed.
Expected total leaf dry mass for each individual was estimated by averaging the dry mass of the collected leaves within a species and multiplying that number by the summed Iax, Rax and G meristem counts for the individual (totalling the expected number of leaves the plant had produced during the experiment). Total vegetative (aboveground) dry mass for each individual was then calculated as the summed dry mass of the main stem(s), plus all side branches, plus the expected total leaf dry mass.
Leafing intensity was calculated by dividing the individual’s expected total number of leaves by its combined dry mass of the main stem and side branches.
Potential fecundity for each harvested plant was estimated by counting the number of seeds in an individual mature fruit for each of five randomly selected fruits. This number was then multiplied by the summed number of collected fruits, counted flowers and missing fruits for each individual.
Branching intensity was calculated by dividing the total number of axillary meristems committed to growth (branching) divided by the combined dry masses of the main stem and side branches.
G (growth) meristem allocation was determined by the total number of axillary meristems committed to growth (branching) relative to the number of axillary meristems committed to reproduction or inactivity, i.e. GIax+RaxGIax+Rax.
R (reproductive) meristem allocation was determined by the total number of reproductive meristems relative to the number of meristems committed to growth or inactivity, i.e. Rax+RapG+Iax+IapRax+RapG+Iax+Iap.
Total bud bank for each individual was determined by the combined total number of axillary meristems committed to growth, inactivity or reproduction.
For each population, the height and number of leaves present prior to initial clipping for both the treatment and control groups were compared using paired sample t-tests (Table 2).
We were interested in exploring potential patterns that may be associated with particular phenologies, life histories and susceptibilities to herbivory. Consequently, rather than applying a general linear model, we performed paired sample t-tests within each population to test for differences in functional traits between treatment and control groups. false discovery rate (FDR) testing was performed within each family of tests (i.e. within the fecundity-, biomass- and leafing-related variables) to account for the multiple comparison tests conducted. Families for FDR testing were assigned due to comparisons being interpreted within groups of related variables.
Simple linear regression analyses were used to examine between-species relationships for mean single leaf mass, and for mean leafing intensity—both recorded for control plants—versus relative mean values (under treatment versus control conditions) for fecundity, total bud bank and total vegetative dry mass, respectively.
All variables were tested for normality using the Shapiro–Wilk test, and were log transformed as needed, following the assumptions of the paired sample t-test, and linear regression analyses. All statistical analyses were performed using RStudio (R Core Team 2015, Version 1.1.456).