Geoscience Reference
In-Depth Information
munities over time until a climax stage was reached, with communities functioning
in similar ways to discrete organisms or 'superorganisms'. This was an attractive
model as it was simple and implied predictability of ecosystem dynamics. Tansley
(1935), who coined the term 'ecosystem', refuted that communities acted as organ-
isms but stated that ecosystems consisted of biotic and abiotic components that
develop following disturbance to obtain a stable dynamic equilibrium. Both of these
early viewpoints assumed that equilibrium states and predictability were inherent
to ecosystems.
This meant that in theory it should be possible to create mathematical models
that would explain and predict ecosystem dynamics and could therefore be used for
ecosystem resource management. Early models, however, had poor predictive capac-
ity, reinforcing the idea that ecosystem modelling and management should focus
only on the very few ecosystem components directly relevant to the managing body
(e.g., Gaichas, 2008). Despite increasing evidence of complexity and variability in
ecosystem dynamics, particularly relating to seral changes in communities, the
mechanistic approach (i.e., nature as machine) to ecosystem processes and concepts
of equilibrium and predictability greatly infl uenced ecosystem modelling and resource
management until the advent of chaos theory and investigations into complexity
and complex systems (Pahl-Wostl, 1995). The legacy of this original focus on stabi-
lity and equilibrium is still being felt today in resource management.
As the 20th century advanced, increasing evidence of individualistic and stochas-
tic mechanisms operating within ecosystems made it clear that ecosystems were
indeed more random, variable and unpredictable than originally supposed over all
spatial and temporal scales (e.g., Pahl-Wostl, 1995). Recognition of stochasticity
and ecosystemic variability across spatial and temporal scales raised important ques-
tions about the possibility of predicting maximum sustainable resource yields and
consequently about the levels of uncertainty and acceptable risk in ecosystem
resource management. This was further supported by an acknowledgement of the
limitations and the poor predictive potential of resource models (e.g., Batchelor
et al., 2002), which often ignored wider ecosystem variability and so compromised
their accuracy. In part, this was due to the reductionist approach of monitoring and
modelling components and processes in isolation, in attempts to understand their
dynamics, rather than adopting a holistic approach for whole ecosystem under-
standing (Pahl-Wostl, 1995). By the 1990s, the concept of ecosystems as complex
non-linear systems was well established (e.g., Kay 2000).
Towards Whole Ecosystem Management for
Resources and Ecosystem Services
At the same time that the inherent variability of ecosystem processes was being
elucidated, the detrimental impacts of poor ecosystem management were also
becoming apparent (e.g., Linton 1970; Darge and Kneese, 1980). In particular,
changing land-use patterns and the intensifi cation of single resource production
(e.g., the clearing of Amazonian tropical forest for ranching) reduced biodiversity
and affected the integrity, resilience and stability of ecosystems, potentially leading
to local ecosystem collapse and, globally, the threat of a human-induced mass extinc-
tion event (e.g., Dale et al., 1994; cf van Loon, 2003). This, combined with the
high-profi le collapse of fi sh stocks, extinction or decline of charismatic species, and
increased risks to human health or living standards meant that conservation and
