NEW CONCEPTS AND METHODS TO DESCRIBE,
MODEL AND QUANTIFY ECOSYSTEM SERVICES ON
DIFFERENT AND MULTIPLE SPATIAL AND TEMPORAL SCALES (Part 2)
Rashed Sheriff
rashed.sheriff@gmail.com
RUBICODE
The wellbeing of humans is integrally linked to the wellbeing of the other species with
which we share the planet. There is now wide acceptance that if the current rate of loss
of biological resources is continued, the result will be catastrophic for humankind within
a few generations. This is a potential threat of immense proportions. The translation of
this threat to biodiversity into tangible and quantifiable factors for use by policy-makers
in decision-making processes is challenging, but vital for conservation. RUBICODE will
contribute to solving this by examining what biodiversity does for us.
Biodiversity provides many services from which humans benefit, such as the provision of
food, fibre and fuel, regulation of air and water quality, flood protection, pollination, control of pests, recreation and ecotourism. RUBICODE will define and evaluate those
components of biodiversity which provide specific services to society. This will increase
our understanding of the value of biodiversity services and, consequently, of the cost of
losing them. This will give decision-makers a more rational base and will help the understanding of the need for adequate conservation policies, which are essential to halting
biodiversity loss.
RUBICODE research themes:
(Rationalising Biodiversity Conservation in Dynamic Ecosystems)

Frameworks and concepts for the assessment of ecosystem services in terrestrial
and freshwater ecosystems.

Approaches for linking ecosystem service provision to
functional traits.

Indicators for monitoring ecosystem services.

Socio-economic and environmental drivers of
biodiversity change.

Strategies for conserving and managing biodiversity and
the services it provides that take account of drivers of
biodiversity change.

Identification of current gaps in knowledge and future
1
research needs.
Conceptual framework
Identification
Define the ecosystem service:
• identify the ecosystem service beneficiaries
• identify the spatio-temporal scale of service delivery
• identify the ecosystem service providers (ESPs)
Quantification
Quantify the ecosystem service demand:
• determine the net level of demand/need for the service Quantify the service-providing
unit (SPU):
• determine the characteristics of organisms necessary for service provision
• quantify the relationships between SPUs and service supply
• quantify the components of biodiversity that support the SPU
Appraisal
Identify and value potential alternatives for providing the service
Evaluate options:
• compare valuations and examine trade-offs
• determine implications for biodiversity conservation
• determine implications for policy and sustainable livelihoods
Service-providing unit (SPU)
“This can be described as the collection of organisms and their characteristics necessary to deliver a given ecosystem service at the level required by service beneficiaries.”
The Following Should Be Known:

The sections of society that need/use the service.

The level at which it is required.
2

The organisms that provide the service

The characteristics of these organisms required to provide the service at the desired level (SPU).
Frameworks: development

Local level case studies - SPU approach, but need more explicit data on beneficiaries, demand, service providers, etc. to study how changes in SPUs affect service provision and to identify consistent trends across case studies that can be
applied broadly.

Regional planning frameworks – trait-based approach, but need more data on
links between species traits and ecosystem service provision to develop maps of
where services are likely to be provided based on species’ distribution maps.

Broadscale mapping studies - need to be extended and refined to account of location of service generation, service beneficiaries and their needs/demands, service flows and future pressures.
AN ECOHYDROLOGICAL FRAMEWORK FOR ANALYZING ECOSYSTEM EFFECTS OF LIVESTOCK PRODUCTION
Introduction
Apart from timber extraction and crop farming, harvesting of biomass from natural ecosystems may also be indirect, through the grazing of domestic animals, mostly in rangelands. An important property of rangelands is their ability to occupy multiple steady
states and therefore different types of biodiversity and different ecosystem functions.
Depending on the external drivers of fires, rainfall, or grazing, rangelands can occupy a
stable configuration dominated by perennial grasses with few shrubs or one dominated
by shrubs with little grass with respective biodiversity levels. Thus, several direct and indirect factors including agricultural trade liberalisation have the potential to push rangelands from a productive and sustainable state to an unproductive state, loss of biodiversity and related ecosystem functions (see Anderies et al. 2002; Perrings and Walker
2004)
Livestock can have a wide range of effects on rangeland ecosystems that can both be
positive or negative and grazing impacts can be seen at the species, community and
landscape levels. Livestock grazing has impacts on individual plants or species, plant
communities, soils and watersheds, and wildlife. In addition, each of these impacts has
secondary or ecosystem-level effects (Miller 2000). Heavy grazing may directly alter
plant species composition and productivity, as well as animal population. It may also indirectly affect invertebrate species, which often play an important role in the maintenance of the stability in ecosystems.
Consequently, livestock management practices impact on rangelands is expected to be
considerable through their influences on the direct and indirect impacts of livestock grazing. Management practices such as fencing, water development, predator control, salt3
ing, vegetation manipulation, and road development can greatly affect biodiversity.
Fencing serves as barriers to movement of wildlife as well as causing injuries to individual animals. Water development for livestock can usurp water used by wildlife, create
areas of overgrazed rangeland, and spread the impact of livestock into new areas.
Predator control can reduce the density of carnivores, which also leads to increased
number of small mammals (rabbits, pikas, voles, mice, etc.). Vegetation manipulation
can decrease wildlife species richness and diversity and lead to invasion by exotic
plants. Burning is often used intentionally in rangeland management to reduce invasion
by shrubs and encourage grass growth. These burning practices are frequently not beneficial to insects or to small mammals and birds (Miller 2000).
The effects of grazing on rangelands also depend on the species of the grazing animals.
For example domestic goats may destroy the plant roots, promoting soil erosion, whereas cattle only eats the above-ground parts of the plants (although, being heavy animals,
they can cause soil erosion indirectly through soil compaction resulting in higher runoff)
(McNeely et al. 1995).
In Southern Africa, different ecological regions influence the predominance of different
kinds of livestock and therefore different pressures on their vegetation's biodiversity. For
example, in South Africa the Karoo region predominantly in the Northern Cape Province
is especially sheep farming dominant while the Bushveld and Lowveld regions in the
Limpopo Province is goat dominated. In other parts such as the Eastern Cape and KwaZulu Natal Provinces with high Karoo and grassland vegetations cattle dominates
(Jooste and van Zyl 1999; Statistics South Africa 2002). Agricultural trade liberalisation
which may change the state of these grazing activities, are accordingly expected to influence the biodiversity in these different regions.
Deacon (1995) has indicated that trade limits are equivalent to a tax on controlled
goods. If such taxes are reduced, this may also imply a reduction in trade limits, and
may lead to increased output in the good. Given less substitutability of input use in the
production of the good, or a market failure in the input market such that it is more or less
a free input, for example a grazing land, it may lead to excessive use of the input with
repercussions for the resource and the environment.
4
In livestock production water is used for:
1. animal drinking needs
2. servicing (e.g. wash & cool animals, clean production facilities, dispose of wastes)
3. processing of livestock products (e.g. slaughter, tanning)
4. production of animal feeds
Main aims of this study – with focus on livestock
1) Include estimates of green water to make feed
2) Explicitly include land use and land cover changes
3) Link discussion of livestock-driven activities
with resulting changes in ecosystem services
5
SCIENTISTS ATTITUDE TO THE ECOSYSTEM SERVICE CONCEPT
AND CONSEQUENCES OF RESEARCH
Page 3
Introduction
As with every new concept about to establish itself as a paradigm, ecosystem services
were greeted with a mixture of applause and scepticism. Judged by the publications on
ecosystem services, five main interests dominate: 1. technical, 2. practical, 3. political, 4.
cynical and 5.economical. Depending on the authors' main interest, research in ecosystem services falls into different groups, aligning nicely with the above five attitudes. In
consequence, scientists communicating about ecosystem services not only mean different things, but they also largely differ in their research recommendations. Recently, several papers have drawn attention to the different definitions; here we want to draw attention to the different research approaches which are derived from the differences in attitudes to the ecosystem service concept.
Why are we interested in Ecosystem Services?
1. It’s a new buzzword, a selling point.
2. ESS raise awareness of Nature’s silent work.
3. As a framework for inter- and trans-disciplinary work.
4. It gives us specific ecosystem properties to focus on.
5. Through monetarisation of ESS we can employ the full economic weaponry to tackle
environmental problems.
What are the implications for science?
1. None: keep doing good research.
2. Showcase research needed.
3. Communication on aims, definitions, scales.
4. Models have to be tailored to specified ESS.
5. Value of non- marketable goods; discounting; defining weights in multi- criteria optimization.
What are foreseeable problems?
1. “Misusing” a buzzword deflates its value.
2. Showcase may turn out counterproductive.
3. Communication is very time-intensive.
4. Tailored models may be inappropriate (wrong scale).
5. Economic forecasts of political steering are very poor.
6
What is the aim of “ecosystem services”?
The MDCSD:
1. Multi-criteria
2. Dynamic
3. Context-specific
4. System Description
SPATIAL PREDICTION OF ECOSYSTEM SERVICES BY MEANS OF
CHEAP AND EASILY AVAILABLE GEODATA
Introduction
Sustainable land management and land use planning require reliable information about
the distribution of those parameters affecting both, landscape processes and services. In
order to provide adequate soil information for the modeling of landscape processes,
such as water movement, nutrient leaching or plant growth, this paper investigates to
what extend cheap and readily available ancillary information derived from digital elevation models and/or remote sensing data can be helpful to predict the spatial patterns of
water retention characteristics of soils and to delineate areas with differing water retention capacity. In our study we focus on paddy soil landscapes in the southeast of China.
The decline of water quality and the increase of water scarcity are the major concerns
for sustainable land management in water-consumptive irrigated rice landscapes. Thus
water saving strategies have to be developed that consequently restrict vertical water
percolation, solute leaching and lateral seepage. However, the establishment of soil and
water protection strategies depends on the knowledge of the spatial distribution of the
factors governing the movement of water and solutes on the one hand and on the
knowledge of the retention characteristics in a particular landscape on the other. In order
to describe the water retention characteristics of rice growing landscapes we used the
available water capacity of the puddle topsoil layer and the saturated hydraulic conductivity of the compacted plough pan as indicators. Both were estimated by using pedotransfer functions (PTF) that consider soil properties (as clay and sand content, bulk
density, horizon depths, organic C-content) as input variables. The latter ones were spatially predicted by applying a regression-kriging model that integrates as well point sampling soil data as digital terrain and/ or remote sensing data. The results of our study reveal that the NDVI (calculated from 29 ASTER (Terra), SPOT, LANDSAT TM and
ETM+) corresponds quite well to the retention characteristics as topsoil water capacity
and the hydraulic conductivity of the plough pan and tends to be indicative of different
levels of water retention in this kind of landscapes. Like other researchers before, we
also found out, that even the incorporation of rather weakly correlated co-variables (e.g.
terrain attributes) into predictive soil-landscape mapping models such as regressionkriging models tends to be suitable to improve the prediction performance.
7
Objectives
1. Identification of spatial dependencies between soil properties (sand, silt, clay, soil organic matter, and horizon depths) terrain attributes and remote sensing data on the
landscape scale.
2. Characterization of spatial correlation structures concerning the elementary soil properties and pedohydrological properties.
3. Comparison of selected prediction methods in order to determine the most appropriate technique to spatially transfer soil data from a limited number of sampling points
to un-sampled locations.
4. Derivation of the spatial patterns of soil property distribution and (soil) water retention
characteristics (soil texture, total C content, puddled layer depth, plough pan thickness, plant available water capacity and hydraulic conductivity).
Research Strategy
From point data to spatial information
8
Prediction of soil property distribution by using aregression krigingapproach
(cf. ODEH et al., 1995)
9
Conclusions
1. The incorporation of ancillary terrain variables can distinctly improve the prediction of
soil properties on the landscape scale even in strongly altered landscapes such as
paddy soil landscapes.
2. The use of terrain variables in a regression kriging model is a suitable method to increase the accuracy of prediction of soil properties and soil characteristics that are
dependent on surface related processes.
3. “Relative elevation” and “wetness index” accounted for 25% of the variation of the Ahorizon’s silt contents.
4. The NDVI calculated for the studied paddy soil landscape tends to be indicative of differing retention characteristics, such as the hydraulic conductivity of the plough pan
and the plant available water capacity of the non-submerged A-horizon.
10
5. The NDVI accounts for 23% of the variability of the topsoil’s AWC and for 24% of the
plough pan’s Ks
6. The highest NDVI values point to locations with always wet or very moist soil conditions and to locations with the best water supply and the highest grain yield.
7. The NDVI is assumed to indirectly reflect the spatial patterns of C and N distribution.
The NDVI explained up to 29% of the variation of the total C content, 26% of the total
N content.
8. The retention characteristics of paddy soils depend on the age of the paddy fields.
9. The degree of structured spatial correlation of the soil variables under study depends
on the spatial scale and on the land use history.
GLOBAL CHANGE AND HUMAN CHANGE: NATURAL ECOSYSTEM
DESIGN AND CONTROL IMPERATIVES NATURAL ECOSYSTEM DESIGN AND CONTROL IMPERATIVES FOR SUSTAINABLE ECOSYSTEM
SERVICES FOR SUSTAINABLE ECOSYSTEM SERVICES
Husbanding for Ecosystem Services
The Near-Steady State Imperative
1.
·
2.
Services Good conservation practices are probably “good” because they promote
near-steady state dynamics. They . . .avoid large changes favor small changes
reduce disturbance impacts promote stability enhance predictability engender
well-behavedness reduce exogenous subsidies lessen maintenance costs minimize surprise conserve and enhance biodiversity encourage sustain ability evidence linear dynamics etc.
Examples: organic & anthroposophic farming stem-based and shelter wood forestry ecosystem-based population management integrated living resources
management etc
Production-intensive practices are probably “less good” because they induce and
sustain deflections from steady state. They: cause large changes do not encourage small changes increase disturbance impacts decrease stability reduce predictability induce poorly-behaved dynamics reduce need for subsidies decrease
maintenance costs surprise happens diminish biodiversity discourage sustainability behave nonlinearly etc.
11
·
Examples: industrial & monoculture agriculture industrial & monoculture forestry
industrial fishing & aquaculture MSY-based fish and wildlife management etc.
The Nonlocality Imperative
Indirect/direct effects ratios increase in proportion to the following network properties:
·
number of compartments
·
density of adjacent connections
·
standing stocks in compartments
·
cycling (non-feedback and feedback)
·
strength of adjacent (“direct”) flows
In extracting ecosystem goods and services, these network properties are to be manipulated to increase or decrease the relative strength of indirect effects as appropriate to
the applied problem at hand
Typically, “more” of the above characteristics (biodiversity, connectivity, stocks, cycling,
and interaction strengths) tend to materialize in natural ecosystem development
Nonzero Imperative
Ecosystem Services. The United Nations Millenium Ecosystem Assessment
(2005; www.milleniumassessment.org) comprehensively reviewed the status and
change of Supporting Services, such as nutrient cycling, soil formation, and primary production, provided to humankind by the natural world. Ecosystem services were classified
into three categories. Provisioning Services refer to resource supply—food, fiber, water,
fuel, and other needed materials and energy. Regulation Services are associated with
climate, floods, disease, water quality, and other factors in control of provisioning.
Cultural Services include aesthetic, spiritual, educational, and recreational aspects of
ecosystems. These three categories—material, cybernetic, and virtual, respectively—are
all very different, and thus pose a challenge for unification within a single theory. Of
course, it is not a given that theory is absolutely necessary for an applied science of
Ecosystem Services to be successful. Yet, such theory is already manifested, implicitly,
in the elements of geosphere, atmosphere, hydrosphere, biosphere, and noosphere woven together in the world's ecosystems. The scientific premise has to be that the acquisi12
tion of ecosystem services will be more correct and powerful if it follows from theoretical
understanding of the deeper principles underlying man-and-nature coupling. This paper
attempts a preliminary sketch of such a theory from the perspective and methodology of
Network Environ Analysis.
The Nonzero Imperative. Environ analysis begins with the physics and works towards
biology. Of physics' four forces—strong and weak nuclear, gravitation, electromagnetic—the first three serve as implicit background for ecosystem services; only electromagnetic is of explicit concern. Electromagnetic force is transferred between any two focal
nodes in a system (compartments, species, etc.) by binary transactions. These are zerosum because the energy and matter directly exchanged are conservative— 30 what is
lost from one member of an interacting pair is exactly gained by the other. Transactions
give rise to binary relations; these are non-conservative, from which follows nonzerosumness—what one element gives up to another may be less or more than the latter receives. Transactions are objective, but in relations subjectivity enters. Beyond life's necessities, "ecosystem services" reside to some extent in the eye of the beholder—that is,
human needs possess subjective as well as objective elements. Sustainably entraining
human change to global change, by matching evolving patterns of ecosystem services
to changing requirements and perceptions of human needs, requires continuous transformation of transactional zero-sumness— the physics—into relational nonzerosumness—the biology. This is the Nonzero Imperative.
Ecosystem Services and Input Environs. Environs are partition subnetworks of a system's interior network of transactions defined by boundary inputs and outputs. Each input in present time successively drives a set of future interior processes reachable from
its point of introduction. The subnetwork generated as the driving wave propagates, and
ultimately dissipates as boundary outputs, is a future-oriented ("after input") output environ. Similarly, each output results from a set of interior processes in past time reaching
forward from their points of input to the point of exit in present time. This subnetwork is a
past-oriented ("before output") input environ. Ecosystem services in the network context
represent boundary outputs from a natural system to a coupled human user set. A particular service may have multiple origins, specific or diffuse, deep inside the system, ultimately traceable back to points of input. The service as output defines a unique input
environ within the system interior that alone is responsible for generating it. Nothing beyond the borders of this subnetwork contributes anything to generation of the service in
question. Thus, with every ecosystem service is associated a unique input environ containing every subpart of the total system network involved in producing and transporting
that particular service to the system boundary.
Utility and Network Synergism. Ecosystem services convey utility—positive benefits to
the human user when present in ample supply, but negative consequences if limiting.
Signed weighting of transactions and relations to reflect benefits (+) and costs (–) is formulated in the utility portion of network environ analysis. In this, every pairwise transaction and relation within a system, both proximate (direct) and ultimate (indirect), receives
an ordered pair of signs taken from the set {+, –, 0}, and an associated quantitative
13
weighting. The sign pairs by themselves define the kind of interaction, and the weighting
factors the magnitudes of associated benefits or costs. A universal property arising in
utility analysis is network synergism—the tendency for nonzero-sum relations between
interactive pairs of components to become quantitatively more positive than the zero31
sum transactions that (may or may not) directly connect them. This opens the door to
the possibility that humans could learn to improve the interior designs of ecosystems,
and their boundary coupling, and so enhance the outflow of ecosystem services.
The Search for Nonzero. In the theoretical terms of this paper, designing human organizations, values, economies, fashions, and behaviors in a "search for nonzero" aimed at
maximizing network synergism, should become the ongoing work of each generation of
humanity. Having sure knowledge of the scientific principles involved, drawn from a
comprehensive unified theory, could only help in realizing the Nonzero Imperative.
Design with Nature
1. Near-Steady-State Imperative
Natural ecosystems operate near, but not at, or far from, steady states or extrema; dynamics of ecosystems for human purposes should be similarly maintained.
2. Nonlocal Imperative
Indirect effects dominate determination in natural ecosystem networks; depending on
the requirements of particular services, ecosystem designs for human benefits should
either maximize or minimize nonlocal relationships.
3. Nonzero Imperative
Natural ecosystems enhance positive relationships among their constituents; ecosystems for human services should be managed to maximize their expression of mutualistic
and synergistic properties of networks.
Control like Nature
4. Multifactorial Control Imperative
Natural ecosystems are regulated by distributed checks and balances involving many
control variables; extraction of human services from ecosystems should similarly be
managed not by single variables, but by coordinated use of as many control variables as
possible.
14
USING REMOTE SENSING TECHNOLOGY TO ASSESS SPATIAL AND
TEMPORAL VARIABILITY IN THE CONDITION OF AUSTRALIAN SAVANNA LANDSCAPES
Introduction
Unlike Australia’s temperate ecosystems, the tropical savanna landscapes are in relatively unmodified condition (Woinarski et al., 2007). Disturbances such as fire, logging/thinning, cycles of drought and rainfall excess, and grazing occur within these landscapes and can be both temporally and spatially variable, increasing landscape heterogeneity but rarely resulting in well-defined patch boundaries (Pearson, 2002; Woinarski
et al., 2005). Instead, the tree-grass structure varies continuously at local to landscapescales (100s-1000s ha).
Traditional models for analysis of heterogeneity in landscape structure and relevant
functionality for fauna species have depicted landscapes as static arrays of vegetation
patches with discrete boundaries between habitat and matrix patches (Forman and Godron, 1986; Wiens, 1976). However, many rangeland and savanna landscapes with
gradual spatial and temporal variation in structure do not fit this model.
Many species require more than one type of habitat during different life stages or for different functions such as foraging or nesting (Fahrig, 2003; Law and Dickman, 1998;
Wiens, 1997). Thus, to gain an understanding of change in ecological function with
change in landscape structure it is important to recognise the ecological value of a variety of landscape components, and their changes over time. Few previous studies have
considered these varying requirements.
This work aims to provide quantitative and continuous measures of habitat quality of relevance to fauna in a tropical savanna landscape. In addition this work will investigate
how these measures of habitat quality vary over time by looking at time series of remote
sensing imagery, and measure how important temporal changes are for biodiversity.
15
The fragmentation model

A dominant paradigm in landscape ecology
- discrete boundaries
- patch and matrix
- boundaries defined, often subjectively, prior to analysis

applies well to landscapes subject to intense human modification

particularly extensively cleared agricultural or urbanized landscapes

valuable insights into the ecological impacts of humaninduced habitat destruction

quantifiable, easy to calculate metrics and statistics for landscape analysis
through readily available.

structural variation occurs at a range of spatial and temporal scales
16

The fragmentation model fails to capture fine-scale and within patch heterogeneity

may not fit well to landscapes that do not exhibit discrete boundaries between
vegetation cover types
Future Directions
Approach is relatively basic and exploratory. More sophisticated spatially continuous
models of landscapes could be developed from satellite imagery differentiating woody
and non-woody vegetation cover, including bare ground indices etc. These models could
then be related to behaviour of individual species or functional groups.
Climatic variation is of great importance in explaining in temporal variability in diversity
and this could be incorporated into the models and the relative importance of spatial
versus temporal variation could be further investigated by including variables explaining
spatial heterogeneity at a variety of extents
Results & Discussion
This work establishes a quantitative relationship between continuous spatial variation of
habitat elements and the diversity and abundance of fauna, and investigates temporal
changes in those relationships. Preliminary results from one time step results show that
there is much variation in the response of individual species to the variation in spatial
structure of different habitat elements. Different species respond to different vegetation
elements depending on their behavioural and foraging ecology, and general predictions
about species diversity and abundance need to consider landscape gradients, as well as
patch and site factors. Weighting and combining the habitat elements by relative importance results in a continuous surface of overall habitat quality. This model remains
ecologically relevant as the new measures can explain a significant proportion of spatial
variation in abundance of individual species and diversity of species. Temporal variation
in landscape function, which is often ignored due to time and resource constraints, is also of importance.
Using remote sensing technology and time series data allowed us to describe how landscape function can vary across a spatial continuum and through time. The approach
used in this study was ecologically relevant as it takes into account the different habitat
requirements of a variety of species, instead of assuming a generic response to landscape structure covering multiple species. This study also takes into account temporal
variation in landscape structure and function which is a common shortcoming of most
studies of the influence of landscape structure and composition on fauna. Of particular
relevance to savanna ecosystems is the ability to quantitatively measure continuums in
habitat structure since these landscapes rarely exhibit distinct boundaries between
vegetation cover types, nor are the fauna limited to one vegetation type.
References
·
http://www.ecology.uni-kiel.de/salzau2008/
17
·
http://www.rubicode.net/
·
http://www.nbu.ac.uk/biota/rubicode_page.htm
·
http://www.southwestnrm.org.au/information/downloads/PriceB_paper.pdf
·
http://findarticles.com/
·
Dormann, C.F., Gruber, B., Haase, D. & Seppelt, R. (2008). Scientists’ attitudes to the ecosystem
service concept and consequences for research. Workshop at Salzau Castle, - ECOSYSTEM SERVICES Solution for problems or a problem that needs solution?, Kiel, Germany.
·
Anderies et al. 2002; Perrings and Walker 2004
Part 1
18