Tropical Forest Community Ecology

Constraints on Succession 371

with one suite of life-history traits early in suc-
cession versus different life-history traits later on.
For both of these conceptual models of succession,
species turnover occurred due to known correla-
tions amon gplant size, lon gevity, and slow growth
(Table 22.1). Thus, succession was a population
process that could be understood only by knowing
the interspecific differences in species life-history
traits.Inaddition,Pickett(1976)arguedthatearly
occupants would inhibit the establishment and
growth of later occupants. These two gradient-in-
time models explicitly identified dispersal ability
as a critical life-history attribute of pioneers. For
example, Pickett (1976) noted that long-range
dispersal would be required to reach the interior
of large disturbed patches. Unfortunately these
approaches failed to incorporate specific mech-
anisms of resource competition that permitted
residents to resist displacement or what factors
caused residents to decline in abundance other
than shorter lifespans.

Interaction categories

In a compellin greview of previous work and
theory, Connell and Slatyer (1977) described how
succession might occur via three categories of
interaction: facilitation, tolerance, and inhibition
(Table 22.1). They pointed out that the inter-
action amon gearlier and later species could be
positive (facilitation), neutral (tolerance), or neg-
ative (inhibition). They highlighted how some
existin gmodels focused on resident speciesinhibit-
inglater successional species (e.g., pre-emptive
initial floristics) while some focused on resi-
dent speciesfacilitatinglater successional species
(e.g., the nucleation model). Pickettet al. (1987)
pointed out several problems with these models,
notin gthat they were actually broad cate gories
of mechanisms (e.g., inhibition could occur via
resource competition or allelopathy) that underlie
species change and that each model could oper-
ate simultaneously durin gany succession even
at the same site. Also, the interaction between
any given pair of species could be both facilita-
tive and inhibitory. Nonetheless, if interactions
could be summed amon gnumerous pairs of
species, these three categories would allow succes-
sional interactions to be classified jointly as either

positive via facilitation or negative via inhibition

or neutral. The interaction categories of Connell

and Slatyer (1977) did not explicitly consider

space or impacts of herbivores, although herbi-

vores were briefly considered. Connell and Slatyer

did point out that wide dispersal and numerous

propagules would cause pioneer species to be the

earliest dominants but subsequently focused on

successional dynamics thereafter. They concluded

that inhibition by earlier dominants predomi-

nated durin gsecondary succession. Thus they

suggested that longevity, stress tolerance, and low

resource tolerance should be crucial life-history

traits amon glater successional species.

Vital attributes

Noble and Slatyer (1980) developed the vital

attributes model whereby particular disturbances

produce conditions that can be best exploited by

species with appropriate life-history traits. Species

were classified accordin gto three key traits: mode

of dispersal, ability to regenerate after distur-

bance, and timin gof reproduction and senes-

cence. Like the work of Pickett (1976) and Drury

and Nisbet (1973) this model took a population-

based approach focusin gon the life-history traits

that led to early arrival at a site and those that

led to later arrival and allowed for long-term per-

sistence. Dispersal ability was granted a more

prominent role in influencin gspecies availabil-

ity than in the gradient-in-time models. Once

again, however, the model was not explicitly spa-

tial and the actual underlyin gmechanisms that

caused species displacement and turnover were

not explicitly considered (Table 22.1). This model

did not consider enemies, and competitive inter-

action amon gplant species was de-emphasized

relative to the other models.

Resource ratios

Tilman (1985, 1988) developed the resource ratio

model of succession whereby species turnover

durin gsuccession is driven by interspecific com-

petition that depends upon the shift from high

light/low nutrient conditions early in succes-

sion to low light/high nutrient conditions later

in succession. Tilman (1985) concluded that