Discussion
Globally, canopy-dominated marine ecosystems are decreasing, shifting to
“urchin barrens” dominated by crustose algae and herbivores (Ling et
al. 2015), or turf-dominated ecosystems (Filbee-Dexter & Wernberg
2018). When canopies are lost, marine forests decline in productivity
(Chapman 1981), habitat complexity (Miller et al. 2015), and diversity
(Graham 2004). Given the ecological and economic importance of these
systems, much attention is being paid to factors that might contribute
to their decline or prevent their restoration (Gorman & Connell 2009).
In particular, crustose and turf-forming algae are thought to be major
barriers to the persistence of marine forests, especially under climate
change (Connell & Russell 2010), given the assumption that these turf
and crustose taxa compete with and exclude canopy-forming species
(Connell et al. 2013).
We tested the widespread hypothesis that turf species in marine forests
have negative effects on canopies (O’Brien & Scheibling 2018). We found
that, although turf algae do compete with the canopy, competition occurs
in a limited number of cases and the effect of the turfs on the canopy
was highly variable, often neutral, and even facilitative in many cases
(Table 1). Importantly, we found that in subtidal ecosystems, turf
species do compete with the canopy (primarily kelps; Fig. 4), which
supports the existing conceptualization of competition as a driving
mechanism in subtidal kelp forests (Dayton et al. 1984; Reed & Foster
1984; Connell & Russell 2010; Connell et al. 2013). Further,
experimental findings thought to contradict the competition paradigm
(e.g., Barner et al. 2016) can be situated in a continuum of
interactions that is predictable from the SGH (Bertness & Callaway
1994; Bennett et al. 2015): negative subtidal interactions shift to
positive, facilitative interactions in the intertidal. The mechanisms
underlying the turf-canopy interactions in this study were variable
(Table 1). It is beyond the scope of this study to describe these
mechanisms in detail (recently reviewed in Edwards & Connell 2012;
O’Brien & Scheibling 2018), but despite the differences in the
physiological, chemical, and physical processes that generate
turf-canopy interactions, generalities emerged in the strength and sign
of interactions.