Landscape moderation
of biodiversity patterns
and processes Teja Tscharntke
Agroecology, University of Göttingen, Germany The Ninth Nekudat Hen Symposium
25 October 2011, Ramat Hanadiv Visitors Pavillon
Landscape moderation
of biodiversity patterns
and processes Teja Tscharntke
Agroecology, University of Göttingen, Germany The Ninth Nekudat Hen Symposium
25 October 2011, Ramat Hanadiv Visitors Pavillon
Focus on hypotheses how landscape context shapes local biodiversity patterns and processes
Empirical evidence provided from studies of our group – mostly on pollination and biological control
Agriculturally driven Global Change Tscharntke et al. 2007,
in: Stewart et al., CABI Publ
Agriculture = key driver
Local and landscape scale intensification
Losses in functional biodiversity
Not all species equally affected
For example, herbivores vs. Predators - biological control, pollination
How on earth can agriculture contribute
to biodiversity conservation? (1) Agricultural land = 40% of terrestrial area
(protected reserves = 12%) (Perfecto & Vandermeer 2010, PNAS)
(2) Spillover across managed and natural systems mitigating extinction risks (species as ‘mobile links‘) (Rand et al. 2006, Ecol Letters; Bianchi et al. 2006, Proc Roy Soc London B)
(3) Agroecosystems can support high biodiversity (Clough et al. 2011, PNAS; Steffan-Dewenter & Tscharntke 2002, Biol Conserv)
(4) Agroecosystems provide more food resources
(Westphal et al 2003, Ecol Letters; Holzschuh et al. 2011, Proc Roy Soc B)
(5) Agroecosystem functioning needs biodiversity (Cardinale et al. 2006, Nature; Scherber et al. 2010, Nature)
Given that agriculture is the main driver of biodiversity loss: How on earth can it also contribute…?
(1) 12-43% of all species live outside reserves (Chappell & LaValle 2011) – and many more matrix-dependent
(2) In human-dominated landscapes, organisms spillover across managed and natural systems ,
Populations often use resources in multiple habitats: linking different habitat types, thereby reducing extinction risk.
(3) Traditional and diverse coffee agroforestry can be extremely rich (“bird-friendly, classic papers) also calcareous grasslands in Central Europe
(4) Agroecosystems are often highly disturbed and simplified, but usually provide large food resources used by non-agricultural species: e.g. pollen used by bees coming from adjacent habitats (or fruits harvested by vertebrates)
(5) Agroecosystems need services to be resilient, which seems to be sometimes forgotten by advocates of the land-sparing concept
– next slide & discussed later
Spatial scale matters Plant Field Landscape
Patterns and processes can look very different depending of spatial scale considered
For example, biocontrol agents or pollinators may respond positively density dependent to their respective resources on a local scale,
but negatively density dependent on a landscape scale
Local scale = experimental approaches; landscape scale = generalization
20 landscape sectors around the city
of Göttingen were digitally mapped (GIS)
We digitized the agricultural landscapes around Gottingen
-- Yellow = arable land
20 landscape sectors around the city
of Göttingen were digitally mapped (GIS) Plots
in the centre
of a circle
-- Research plots (potted plants, arable field, grassland) in the centre of a digitited landscape circle
Simple landscapes (>90% annual crops)
Complex landscapes (>50% noncrop area)
We selected a gradient in landscape complexity
-- asked whether populations and communities in the centre of landscapes were influenced by the surroundings.
-- whether by the directly adjacent surrounding or larger circles.
-- for example solitary wild bees with small ranges, whereas large bumble bees and honey experience their surrounding at a lartger scale.
Martin Schmidt
et al. 2008, J Biogeogr Species-specific functional spatial scales Spatial scale = radius of landscape circle (m)
(best relations of % noncrop area to local spider density) Responses to landscape complexity Arable spiders
In wheat fields
Amazing range of different responses to the surrounding landscapes
Annual field: ground checking to update the many digital maps = huge amount of computer work.
Local biodiversity
is largely influenced by
the landscape-wide species pool. The landscape species pool hypothesis Tscharntke et al. submitted
This is known for a long time, i.e. the paradoxon that key drivers of an ecosystem are outside this system
The landscape species pool hypothesis Bees BUT:
Richness of parasitoids of cereal aphids
in simple as high as in complex landscapes
Ines Vollhardt et al. 2008, Agric Ecosyst Environm Tscharntke et al. 2005, Ecol Letters Landscape complexity,
landscape-wide species pool
and local species richness are usually correlated
Two counterbalancing explanations: more alternative resources in complex landscapes, but higher productivity of aphids and parasitoids in simple landscapes
Dissimilarity of local communities
determines landscape-wide biodiversity
much more than
the fragmentation level of habitat The dominance of beta diversity hypothesis Tscharntke et al. submitted
Steffan-Dewenter & Tscharntke 2000, Ecol Letters Butterflies on
calcareous grasslands The dominance of beta diversity hypothesis
Landscape scale: Theory of Island Biogeography (>40 years old), OR/BETTER: species-area relationships
How landscape context affects species depend on their traits
Here: Higher vs. lower trophic levels (herb vs. plants)
Higher trophic level only more affected when specialised (theoretically based by Bob Holt’s models)
Habitat fragmentation per se (not habitat loss) has been overestimated as driver of landscape-wide species loss
Steep increase of richness with area in the beginning, but levels off soon.
The SLOSS debate
Many small habitats
capture more heterogeneity
determining biodiversity
Protection of a complex of both large and small fragments Tscharntke et al. 2002, Ecol Appl
(33 grassland fragments, 61 butterfly species) The dominance of beta diversity hypothesis
However - SAR does not mean that small habitats are of minor importance
Having many small fragments can enhance biodiversity – as shown here graphically.
SLOSS seems to contrast with expectations from species-area relationships.
Community turnover/dissimilarity can be more important than fragmentation effects.
Also: endangered species
The dominance of beta diversity hypothesis In real landscapes of a given extent:
having many fragments = usually spread over larger distances
Larger distances – more heterogeneity – different community structure (dissimilarity)
In addition: different surroundings - different spillover (colonization)
The dominance of beta diversity hypothesis Experimental exposure of standardized nests
for bees, wasps and their parasitoids
(Tscharntke et al. 1998, J Appl Ecol)
Many studies with this experimental system allowing to analyze species richness of bees and wasps - and of the role of their parasitoids
Jason Tylianakis et al. 2005, Ecology & 2006, Ecology & 2007, Nature Gradient of increasing land-use intensification in coastal Ecuador Rice Pasture Coffee Ab. coffee Forest Highly replicated study
48 quantitative host-parasitoid food webs The dominance of beta diversity hypothesis
Tylianakis et al. 2005, Ecology 48 sites
12 months
9 traps per site
and month
Land-use intensification gradient in coastal Ecuador
Additive partitioning of diversity The dominance of beta diversity hypothesis
Here a short impression how the changing community structure in space and time drives overall biodiversity
(this true in arable weeds as well as in pest-enemy systems)
Low local richness, changing structure in time (in the tropics more), always high dissimilarity between sites (and even more between regions)
Alpha diversity (9 nests/site)
Beta diversity (12 months)
monthly turnover Jason Tylianakis et al. 2005, Ecology The dominance of beta diversity hypothesis Land-use intensification gradient in coastal Ecuador
Additive partitioning of cavity-nesting bee and wasp diversity
In addition to the fact that beta diversity makes up most of the diversity on larger scales, e.g. the landscape
Contrasting pattern of alpha vs. beta diversity
Higher dissimilarity in natural habitat – but need not be the case: heterogeneity of farming practices can also cause great turnover
In conclusion: beta diversity is quantitatively dominant and also can result in different patterns
Effectiveness of conservation management
is highest in structurally simple,
rather than in cleared (i.e. extremely simplified)
or in complex landscapes. The intermediate landscape-complexity hypothesis Tscharntke et al. submitted
Comments