The distribution of diversity: challenges and applications
133
Box 5.3 Connectivity in the marine realm – networks of protected areas
In contrast to terrestrial systems, where the dominant paradigm has been to view protected areas
as isolated islands in a sea of inhospitable habitat (Chapter 8 ), the notion of connectivity among
protected areas has received more consideration in the marine context. Many taxa have pelagic life
stages, and species ranges are commonly very large. Networks of protected areas have therefore
been proposed with connectivity in mind (e.g. Sala et al ., 2002 ). The critical questions are how big,
and how far apart, do reserves need to be in order to maintain the connectivity of their constituent
biodiversity?
Early dispersal models treated propagules as passive drifting particles, while currents were modelled by simple fl ow patterns. More recently, complex individual - based and dynamic metapopulation
models have been developed that incorporate realistic currents, fl ow regimes, larval behaviours and
adult spawning strategies (Fogarty & Botsford, 2007 , and other papers in the same issue). One such
study, in the Caribbean, revealed relatively high levels of self - recruitment ( ≈ 12%) and lack of important additional recruitment from beyond 200 – 300 km (Cowen et al ., 2006 ).
Given that relevant magnitudes of dispersal for most reef fi sh are of the order 10 – 100 km, Cowen
et al . ( 2006 ) suggest that passive dispersal is insuffi cient to maintain viable populations. The biogeographical corollary of the model is that distinct regions of population isolation exist. This inference
accords well with phylogeographical data and has important implications for replenishment of individuals within reserves.
Drawing from the world MPA database (Wood, 2007 ), Wood et al . (2008) assess current marine
protected areas in terms of their network characteristics. They report that roughly 35 – 60 per cent
are large enough to be self - seeding for short - dispersing species, while 2,496 MPAs (56.3 per cent
of the world ’ s MPAs) are within 10 – 20 km of another MPA. Considering both the minimum size and
maximum spacing recommendations, and depending on the particular recommendations that are
followed, between 18 and 49 per cent of MPAs can be deemed to be part of a connected network
(See Table B5.3a ).
Of course, such analyses depend on the accuracy of the models and it is unclear the extent to
which patterns translate across biogeographical systems. For example, dispersal distances typically
increase with increasing latitude; therefore, models calibrated in the tropics may not be transferable
to other regions (Laurel & Bradbury, 2006 ).
Table B5.3a Percentage of the world ’ s marine protected areas by number and area that meet both minimum
size and inter - marine protected area distance recommendations made by (a) Halpern & Warner (2003) , (b) Shanks
et al. (2003) and (c) Palumbi (2003) . (Source: Wood et al. , 2008 – their Table 3)
Minimum size (km
2 )
Within 10 – 20 km
b
Within 20 – 150 km
c
%
% by area
%
% by area
> 3.14
b
34.1
54.6
49.1
80.3
> 10
a
19.9
54.4
29.9
80.1
> 12.5
b
18.4
54.4
27.6
80.0
133
Box 5.3 Connectivity in the marine realm – networks of protected areas
In contrast to terrestrial systems, where the dominant paradigm has been to view protected areas
as isolated islands in a sea of inhospitable habitat (Chapter 8 ), the notion of connectivity among
protected areas has received more consideration in the marine context. Many taxa have pelagic life
stages, and species ranges are commonly very large. Networks of protected areas have therefore
been proposed with connectivity in mind (e.g. Sala et al ., 2002 ). The critical questions are how big,
and how far apart, do reserves need to be in order to maintain the connectivity of their constituent
biodiversity?
Early dispersal models treated propagules as passive drifting particles, while currents were modelled by simple fl ow patterns. More recently, complex individual - based and dynamic metapopulation
models have been developed that incorporate realistic currents, fl ow regimes, larval behaviours and
adult spawning strategies (Fogarty & Botsford, 2007 , and other papers in the same issue). One such
study, in the Caribbean, revealed relatively high levels of self - recruitment ( ≈ 12%) and lack of important additional recruitment from beyond 200 – 300 km (Cowen et al ., 2006 ).
Given that relevant magnitudes of dispersal for most reef fi sh are of the order 10 – 100 km, Cowen
et al . ( 2006 ) suggest that passive dispersal is insuffi cient to maintain viable populations. The biogeographical corollary of the model is that distinct regions of population isolation exist. This inference
accords well with phylogeographical data and has important implications for replenishment of individuals within reserves.
Drawing from the world MPA database (Wood, 2007 ), Wood et al . (2008) assess current marine
protected areas in terms of their network characteristics. They report that roughly 35 – 60 per cent
are large enough to be self - seeding for short - dispersing species, while 2,496 MPAs (56.3 per cent
of the world ’ s MPAs) are within 10 – 20 km of another MPA. Considering both the minimum size and
maximum spacing recommendations, and depending on the particular recommendations that are
followed, between 18 and 49 per cent of MPAs can be deemed to be part of a connected network
(See Table B5.3a ).
Of course, such analyses depend on the accuracy of the models and it is unclear the extent to
which patterns translate across biogeographical systems. For example, dispersal distances typically
increase with increasing latitude; therefore, models calibrated in the tropics may not be transferable
to other regions (Laurel & Bradbury, 2006 ).
Table B5.3a Percentage of the world ’ s marine protected areas by number and area that meet both minimum
size and inter - marine protected area distance recommendations made by (a) Halpern & Warner (2003) , (b) Shanks
et al. (2003) and (c) Palumbi (2003) . (Source: Wood et al. , 2008 – their Table 3)
Minimum size (km
2 )
Within 10 – 20 km
b
Within 20 – 150 km
c
%
% by area
%
% by area
> 3.14
b
34.1
54.6
49.1
80.3
> 10
a
19.9
54.4
29.9
80.1
> 12.5
b
18.4
54.4
27.6
80.0