Rapid loss of global biodiversity compels us to understand how natural
systems respond to disturbance. A major factor influencing this process is
how species interact within these systems. Positive species interactions
(facilitation) may be able to buffer disturbance and influence how ecosystems
respond to species loss. An understanding of the circumstances under which
interactions change would not only contribute to many core issues in
contemporary community ecology, but would allow us to apply such
knowledge to inform new and innovative conservation strategies.
Facilitation through bioengineered stress amelioration is predicted to be
particularly important in habitats of high environmental stress (the stress
gradient hypothesis (SGH)). This is because the stress reduction is essential
for certain associated species to survive. Conversely, the same mechanism
of facilitation will not be required under benign conditions. This is the central
tenet of the facilitation-stress hypothesis: that facilitation becomes relatively
more important as abiotic stress increases. As such, this study aimed to
measure the strength of positive interactions in intertidal benthic systems
across environmental stress gradients.
The study focused on two common intertidal species with different
mechanisms of bioengineering: 1) lugworms (Arenicola marina) that
oxygenate the sediment through bioturbation and 2) sand masons (Lanice
conchilega) that stabilise the sediment through construction of biogenic
tubes. Through observational studies, field manipulations and mesocosm
experiments, my research aimed to find whether the SGH was generally
applicable to multiple systems and mechanisms of facilitation.
An initial observational study was undertaken to investigate whether
interactions of adjacent species with A. marina varied between shores of low
and high ambient hypoxic stress (AHS). Lugworm density was found to
positively relate to both the depth of the oxic-anoxic chemocline and
associated species richness at sites of high AHS. At sites of low AHS,
lugworms had no effect. Results suggested that lugworm depressed the
apparent redox potential discontinuity depth (aRPD) and increased species
richness at sites of high AHS though no trends were found with specific
species or within functional groupings.
A second, manipulative study of lugworm was conducted at sites of high and
low AHS to test if patterns of aRPD depth and species richness observed in
the descriptive study were driven by lugworm density. This study attempted
to add to my observational study by establishing causality. Lugworms were
excluded from the sediment and effects compared to procedural controls and
ambient plots. Lugworms significantly reduced sulphide concentrations at
the deepest depths (18 cm) at all high AHS sites, but also at one of the low
AHS sites. In contrast to the observational study, lugworms had no effect on
species richness. However, Corophium spp. were always negatively affected
by lugworm; when present, Bathyporeia spp. always benefited from
lugworms; and Scoloplos armiger showed significant, but highly variable,
responses at a site level. I suggest that there was an insufficient difference
in AHS between sites to detect any differential effects of bioturbation on
species' distributions. Furthermore, we propose that effects of lugworms on
species' densities that were recorded occurred because of mechanisms of
bioengineering other than stress amelioration e.g. funnel and cast formation.
As these interactions did not occur because of stress amelioration, they
would have operated independently of the stress gradient.
Field manipulations were undertaken on L. conchilega to see whether, in line
with the SGH, adjacent species interacting with the tubeworms benefited
from mimics at high, compared to low stress (current speed). Effects of
different densities of tubeworm mimic on an associated infauna community
were studied at relatively higher and lower current speed, modified in situ by
Venturi flumes. Results indicated that L. conchilega tubes increased
sediment shear strength and maintained species richness as current speed
increased possibly as a result of buffered erosion. Small, surface-dwelling
organisms appeared to be promoted preferentially. Evidence suggested that
sediment-stabilising effects of tubeworms supported the general SGH.
Mesocosm studies were used to further investigate how interactions with
L. conchilega mimics shifted over a more complete stress gradient whilst
mitigating against confounding effects. In contrast to my field manipulations,
the mesocosm experiment aimed to show how effects changed across
multiple current speeds in order to find when facilitation manifests and the
shape of the facilitation-stress relationship i.e. whether it is accelerating,
asymptotic, or hump-backed. Effects of increasing current speeds on a
representative community were recorded with and without mimics present.
Interactions were measured as a change in live biomass for each species.
The only species found to significantly benefit was Corophium volutator,
whose mortality was buffered from flow-associated disturbances at current
speeds of 9 cm.s-1. My experiment was limited by not generating high
enough current speeds and I hypothesise that, had I been able to generate
these higher speeds, I would have seen effects on more of the species. This
has implications for how pair-wise facilitation manifests at a community level
at different stress levels.
My findings suggest that facilitation is an idiosyncratic, though important,
process determining how communities respond to changing environments.
They indicate that the basic SGH may be too simple to apply to all natural
systems and I suggest that future research be directed in finding which
systems the SGH applies to in order to refine and develop new conceptual
models that are more representative of real communities and environmental
conditions. As it stands, individual site-specific knowledge is required in
order to use facilitation in conservation and restoration projects.