1. The role of cascades in natural communities has been extensively studied, but interactions between trophic and facilitation cascades are unexplored. In the White Sea (65° N) shallow subtidal bivalve primary facilitators provide hard substrate for secondary facilitator barnacles, that in turn provide substrate for conspecifics, ascidians, red algae, and multiple associated organisms, composing a multi-level facilitation cascade. Previous research revealed that predation by the whelk (Boreotrophon clathratus) accounts for ~7% of adult barnacle mortality. Low whelk abundance limits their effect, with barnacles living on conspecifics several times more vulnerable to predation than those living on primary substrate.

2. Trophic cascades can selectively shield foundation species from consumers, and hence may affect the structure and length of facilitation cascades. We tested the hypothesis that low abundance of the whelks results from mesopredator predation on their juveniles. Depending on the magnitude of the effect, this would mean that a trophic cascade controls the abundance of barnacles on all substrates or only barnacles living on conspecifics. We also suggested that barnacles on primary substrates and conspecifics facilitate different dependent assemblages.

3. We manipulated the presence of crab and shrimp mesopredators in field caging experiments to assess their effect on whelk recruitment. In a field survey we compared the assemblages of sessile macrobenthic organisms associated with barnacles living on different substrates.

4. Caging experiments evidenced that crab and shrimp mesopredators reduce whelk recruitment by 4.6 times. Field data showed that barnacles on primary substrate and on conspecifics promote different dependent assemblages including secondary facilitator ascidians.

5. Although mesopredators do not shield barnacles from elimination, their absence would restrict them from living on conspecifics. Barnacles on conspecifics are functially different from barnacles on primary substrate, and can be concidered a separate level of the facilitation cascade. Trophic cascades thus can generate community-wide effects on facilitation cascades by affecting their structure and possibly length.

Field experiments. We tested mesopredator effects on Boreotrophon clathratus (hereafter 'whelk') recruitment to barnacle clusters in a series of year-long field experiments conducted at a 12 m deep subtidal site in the Solovetsky Islands (65°01.180’N, 35°39.721’E, see Yakovis & Artemieva, 2015). In July 2009-2012 and 2015 we collected empty shells with live Balanus crenatus and similar shells with empty barnacle tests (the latter are almost equally abundant in the field). We defaunated these shells except of adult barnacles (4 or more annual growth rings) and their empty tests and attached them in alternating order to the bottom of 300×375×70 mm plastic cages covered with 2.5 mm nylon mesh (2-3 shells with live barnacles and 2-3 shells with empty tests per cage). Boreotrophon is a direct-developer with crawl-away recruits, attaching egg masses to hard substrates, with a lifespan of several years. The mesh is permeable for juvenile whelks with shell height within 7 mm (Yakovis & Artemieva, 2015).
Each year we deployed a new set of cages (which were collected next year) randomly distributed between 2-5 treatments: (i) full cages (predator exclusions) and (ii) open cages (unmanipulated controls, no mesh, subject to normal predation) in all years; (iii) partial cages to control for caging effects, similar to full cages but with two side windows 175×50 mm each, in all years except 2012; (iv) cages with predatory spider crabs Hyas araneus (crab enclosures) in 2015 only, and (v) cages with predatory shrimp Spirontocaris phippsi (shrimp enclosures) in 2015 only. The set of cages deployed in 2015 was also used to assess the effect of crustacean predators on other mobile fauna associated with barnacles (Yakovis & Artemieva, 2019). There were 2 cages/treatment/year in 2009-2010, 2010-2011, 2011-2012, and 2012-2013. In 2015-2016 there were 6 cages/treatment, except for predator exclusions, which were 8, and open cages, which were 5. We exposed all the cages anchored to the bottom in a haphazard pattern (≥0.5 m apart) for one year, collected them, counted, weighed and individually measured (since 2011) crabs and shrimp with carapace length ≥2.5 mm and whelks with shell height ≥0.5 mm found inside (except for crustaceans in open cages, which were too mobile to be sampled in absence of mesh). For details of the 2015-2016 predator enclosures experiment see Yakovis & Artemieva (2019).
In the field whelks, crabs and shrimps concentrate in patches formed by barnacles and their empty tests rather than on unstructured sediment. Thus, the abundance of predators is more related to that of adult barnacles rather than bottom area. To account for that, barnacle weight in each cage was estimated in the end of the experiments from aperture length measurements using allometric relationships (Yakovis & Artemieva, 2015). Empty barnacle tests were also measured and their equivalent weight was calculated using the same allometric relationships. The sum of calculated weights of live and dead barnacles is hereafter called equivalent barnacle weight, EBW, 151±6 g per cage (n=53).

Field survey. To compare associated assemblages of sessile organisms between barnacles attached to primary and secondary hard substrates (which are differently affected by predators) we used eight samples obtained by SCUBA divers at the experimental site (65°01.180’N, 35°39.721’E) in July 2012-2014. Each sample contained all the hard substrates visible on the sediment surface collected from a 0.25 m2 square frame haphazardly placed on the bottom (10-26 such primary substrates per frame). We counted and identified all the sessile macrobenthic organisms larger than 0.3 mm by substrate type; the individuals attached to first- and second-layer barnacles were recorded separately.

References.
Yakovis E., & Artemieva, A. (2015). Bored to death: community-wide effect of predation on a foundation species in a low-disturbance arctic subtidal system. PLoS ONE 10:e0132973. http://dx.doi.org/10.1371/journal.pone.0132973
Yakovis, E., & Artemieva, A. (2019). Epibenthic predators control mobile macrofauna associated with a foundation species in a subarctic subtidal community. Ecology and Evolution, 9, 10499– 10512. https://doi.org/10.1002/ece3.5570

Tabs 'Experiments 2009-2012' and 'Experiments 2015-2016' contain raw data from the field experiments. The tab 'Field Survey' contains raw data from the field survey.