<br />Joseph H. Connell
<br />
<br />left that they still search the bottom com-
<br />pletely with their feeding activities and eat
<br />all the eggs of their own species that are
<br />laid in early summer, thus completely
<br />suppressing the next generation. Only
<br />after the two-year-olds emerge can a new
<br />generation of eggs start to develop in the
<br />hypolimnion. In shallow water, growth
<br />and development are faster, all emerge
<br />at the end of one year, and space is left
<br />for a new generation each year. Thus
<br />predation thins out the population in
<br />shallow depths, but in deeper waters harsh
<br />physical conditions (anoxia) exclude pred-
<br />ators for a greater proportion of the year.
<br />The prey apparently survive these harsh
<br />periods so that, with less predation, popu-
<br />lation density is greater, as is competition
<br />for feeding space.
<br />The studies described above have dem-
<br />onstrated the dominant role of predation
<br />in determining the relative abundance
<br />and distribution of species in fresh-water
<br />lakes. The community structure is com-
<br />pletely different, depending upon whether
<br />fish happen to be present or not. If not,
<br />the presence or absence of amphibian or
<br />invertebrate predators produces a differ-
<br />ent community structure. Competition
<br />may occur if the intensity of predation is
<br />so low that high population densities of
<br />herbivores develop.
<br />In the sea. On rocky shores in temper-
<br />ate zones, grazers and predators often
<br />keep their prey populations so low that
<br />competition is prevented. The evidence
<br />for this comes from experiments in which
<br />grazers or predators were removed. The
<br />classic experiment of Jones (1948), who
<br />removed 15,000 limpets from a strip of
<br />rocky shore, demonstrated the great effect
<br />
<br />472
<br />
<br />of these herbivores on abundance of
<br />algae. The algae quickly colonized and
<br />covered the surface for several years.
<br />Others have repeated the experiment
<br />elsewhere with similar results (Southward,
<br />1953, 1964; CastenhoIz, 1961; Haven,
<br />1966, 1973; Dayton, 1970). The same
<br />effect of grazing by sea urchins on algae
<br />in tide pools or rocks below the intertidal
<br />region has been demonstrated by experi-
<br />mental removal (Kitching and Ebling,
<br />1967; Jones and Kain, 1967; Paine and
<br />Vadas, 1969; Dayton, 1970). Lastly, by
<br />excluding herbivorous fish from intertidal
<br />and subtidal areas on coral reefs, Stephen.
<br />son and Searles (1960) and Randall (196 I)
<br />have found that they too keep algae and
<br />sea grasses grazed down. Unplanned ex-
<br />periments on a larger scale have shown
<br />the same effect. Grazers were killed by a
<br />spill of fuel oil from a tanker wreck
<br />(North, Neushal, and Clendenning, 1964),
<br />and by detergents used to clean shores of
<br />crude oil (Smith, 1968). Algae colonized
<br />and grew profusely after the grazers were
<br />gone.
<br />On the middle and lower shore levels
<br />experimental removal of predators has
<br />shown that they also often eliminate ses-
<br />sile animals before these reach maturity.
<br />All barnacles were eaten by predatory
<br />snails within 18 months in Scotland
<br />(Connell, 1961a) or 12 to 15 months in
<br />Washington (Connell, 1970; Dayton,
<br />1971) and New Zealand (Luckens, 1970),
<br />except where predators were excluded by
<br />cages. Mussels survived only in cages
<br />(Connell, unpublished; Dayton, 1971) or
<br />when the predatory starfish were picked
<br />off by hand (Paine, 1966, 1971).
<br />Yet despite this very heavy grazing and
<br />
<br />
<br />16 Producing Structure in Natural
<br />Communities
<br />
<br />predation, some plants and sessile or sed-
<br />entary animals survive to maturity and
<br />live a long time at low shore levels. The
<br />clue to how this happens is in the age
<br />structure of these populations: they often
<br />consist of widely spaced older year-classes.
<br />I have studied in detail the population
<br />dynamics of one such species, the barnacle
<br />Balanus cariosus on San Juan Island,
<br />Washington (Connell, in preparation). At
<br />the start the population consisted of
<br />classes aged 2, 4, and at least 10 years.
<br />Over the next I3 years, snail predators ate
<br />. all the young that arrived every year at
<br />two different study sites, with three excep-
<br />tions: once at one site and once at each
<br />of tWO different shore levels in different
<br />years at the other site. In these instances
<br />some individuals survived to the age of 2
<br />years, at which time they were invulner-
<br />able to all the common predators except
<br />the very large starfish Pisaster ochracells.
<br />I tested the invulnerability to smaller
<br />predators by protecting the barnacles in
<br />cages for varying lengths of time and then
<br />removing the cages or allowing predators
<br />to enter. All B. cariosus younger than 2
<br />years were quickly eaten, whereas only
<br />seldom were older ones eaten. Dayton
<br />(1971) has since confirmed these findings.
<br />As described earlier I have found that,
<br />under natural conditions, B. cariosus sur-
<br />vives to this invulnerable age only occa-
<br />sionally, so that the populations consist of
<br />"dominant year classes." This age struc-
<br />ture is produced by occasional escapes
<br />from the intense predation to which they
<br />are usually subjected.
<br />Other organisms may also reach a size
<br />at which they are invulnerable to preda-
<br />tors. Kitching, Sloane, and Ebling (1959)
<br />
<br />473
<br />
<br />
<br />found that the mussel Mytilus edulis be-
<br />came invulnerable to attack by crabs.
<br />Only one species of very large crab, Can-
<br />cer pagurus, could break open the larger
<br />mussels available. This case is similar to
<br />that of B. cariosus; the prey can reach a
<br />size invulnerable to the smaller species of
<br />predators, but there exists a very large
<br />species of predator to which it never is
<br />invulnerable. Dayton (1971) suggests that
<br />Mylilus californian us may reach a size at
<br />which it becomes invulnerable to Thais;
<br />although this is probable, no evidence
<br />exists, since the predators ate all sizes
<br />offered in the experiments.
<br />Some species of intertidal algae may be-
<br />come invulnerable to grazers. Southward
<br />(1964, Table 2) found that the alga Fucus
<br />vesiculosus, which colonized a shore after
<br />limpets, Patella vulgata, were removed,
<br />maintained a complete cover for the next
<br />2 years. During this period many small
<br />limpets colonized the rock beneath the
<br />canopy of large algae. After 4% years the
<br />limpets were much larger and the algal
<br />cover had been reduced to 22%. The
<br />smaller limpets evidently did not attack
<br />the large algae, feeding instead on the
<br />smaller plants that colonized beneath the
<br />canopy. When they grew larger, they ate
<br />the large algae. It seems reasonable to
<br />draw the conclusion that large algae are
<br />invulnerable to small grazers but not to
<br />large ones.
<br />Grazing and predation are often ex-
<br />tremely intense under the more "benign"
<br />conditions of the lower shore, and "es-
<br />capes" to invulnerable size occur only
<br />rarely. The only data as to the frequency
<br />of such escapes in natural conditions are
<br />those I have given for Balanus cariosus.
<br />
<br />
<br />
<br />
|