Download Introduction Results and implications

Designing effective marine reserve networks for metacommunities
Tarik C. Gouhier
http://www.northeastern.edu/synchrony
Introduction
I use a spatially-explicit version of the classic Rosenzweig-MacArthur
model5 to simulate the dynamics of prey N1 and predator N2 at each site x.
The prey species undergoes logistic growth and the predator has either a
type II (j=1) or a type III functional response (j=2). Dispersal d is modeled
as a gaussian kernel k with constant advection/diffusion rates resulting in
a total extent representing 3% of the spatial domain. Reserve networks
characterized by different levels of protection and spacing are implemented by varying harvest-related mortality hi(x) at each site x:
dN1 (x)
⎛ N (x) ⎞ aN N 2 (x)
= rN1 (x) ⎜ 1 − 1 ⎟ −
− h1 (x)N1 (x) − dN1 (x) + ∫ dN1 (y)k(x − y)dy
⎝
dt
K ⎠ 1 + avN1j (x)
− L/2
L/2
j
1
dN 2 (x) aN (x)N 2 (x)
=
− mN 2 (x) − h2 (x)N 2 (x) − dN 2 (x) + ∫ dN 2 (y)k(x − y)dy
dt
1 + avN1j (x)
− L/2
j
1
L/2
Global abundance
(% change)
0
31
11 6
3
1
90
60
30
5
0
0
31
11 6
3
1
90
60
30
0
(d)
0
40
31
11 6
3
1
90
60
30
0
(e)
0
(f)
31
11 6
3
1
90
60
30
0
0
0
31
11 6
3
1
(g)
Yield
(% change)
The model
The figure shows the change in the mean abundance of the prey (left
column) and the predator (right column) for reserves characterized by different levels of protection and spacing relative to the ‘no reserve scenario’
(horizontal plane) across the entire metacommunity (a, b), within reserves
(c, d), in unprotected areas (e, f), along with the total yield or harvest size
(g, h).
(b)
10
(c)
Abundance
within reserves
(% change)
However, these predictions are based on conceptual models that typically
ignore two features that are ubquitous in marine environments: (1) species
undergo strong spatiotemporal fluctuations in their abundance, especially
when exploited, and (2) species are coupled to each other via competition
and predation. Here, I use a dynamic predator-prey model that has been
validated in a natural marine system to show that realistic spatiotemporal
fluctuations in population abundance have important implications for designing reserve networks that protect entire metacommunities of trophically and spatially coupled species.
Results and implications
Predator
(a)
10
Abundance
outside reserves
(% change)
The proliferation of efficient fishing practices has promoted the depletion
of commercial stocks around the world and caused significant collateral
damage to non-target species and marine habitats. Empirical studies have
shown that marine reserves, areas of the ocean where extractive activities
are either partially or completely prohibited, can play an important role in
reversing these harmful effects by allowing species to rebuild their numbers both within and beyond protected areas1-3. Networks of reserves can
generate even greater benefits by protecting species across a larger portion of their natural range and providing ‘spatial insurance’ against local
catastrophes such as toxic spills3. Conventional wisdom suggests that in
order to be effective, reserve networks must be spaced far enough to provide ‘spatial insurance’ against catastrophes but not so far as to prevent
connectivity in the form of larval subsidies to flow from one protected area
to the next4.
Prey
90
60
30
0
31
(h)
11 6
3
1
90
60
30
0
0
0
31 11 6
3
Distance be
tween
reserves (%
)
1
0
30 )
60
(%
31 11 6
90 tion
3
c
D
is
e
ta
t
n
c
e
betwee
Pro
reserves (% n
)
1
0
30 )
60
%
90 tion (
tec
o
r
P
Reserve networks whose spacing is smaller than or equal to the extent of
dispersal (i.e., distance between reserves of 3%) lead to a mild increase
in predator abundance but minimize prey abundance because of trophic
cascades. However, reserve networks that are much farther apart than the
extent of dispersal (i.e., distance between reserves of 11%) simultaneously maximize the abundance of both the predator and the prey, along with
the yield of the prey5.
The critical reserve distance that maximizes the abundance of both the
prey and the predator corresponds to the extent of spatial autocorrelation
in adult abundance (i.e., the extent of adult patchiness), and only emerges
when realistic population fluctuations are coupled in space via dispersal.
Importantly, the extent of adult patchiness can be readily computed for a
variety of species by applying simple spatial statistics to existing datasets.
Hence, this potentially broadens the applicability of reserve networks to
systems for which connectivity and dispersal rates are either unknown or
poorly documented.
Conclusions
Overall, these results suggest that using the extent of adult patchiness
instead of the extent of larval dispersal as the size and spacing of reserve
networks is critical for designing community-based management strategies. By emphasizing patchiness over dispersal distance, these results
show how incorporating the complex temporal fluctuations in species
abundance observed in nature can actually simplify management guidelines and reduce uncertainty associated with the assessment of dispersal
in marine environments5.
References
1. Roberts et al. (2001). Effects of marine reserves on adjacent fisheries. Science 294
2. Gell & Roberts (2003). Benefits beyond boundaries: the fishery effects of marine reserves. Trends Ecol. Evol. 18
3. Gaines et al. (2010). Designing marine reserve networks for both conservation and fisheries management. PNAS 107
4. Sale et al (2005). Critical science gaps impede use of no-take fishery reserves. Trends Ecol. Evol. 20.
5. Gouhier et al. (2013). Designing effective reserve networks for non-equilibrium metacommunities. Ecol. Appl.