Spatial Ecology

Spatial Ecology

1. What is spatial ecology?

   • Spatial patterns and distribution of plants and animals
   • In relation to their environment: climate, resources, geography

2. Spatial data

   • Detecting spatial patterns and distribution
     • distribution data; presence-absence data; presence only data; movement data
   • Methods
     • GPS collars; transects; habitat mapping; remote sensing

3. Spatial ecology in conservation

   • Overall threats to biodiversity
     • Habitat loss and land use change; hunting; climate change
   • Movement and space use
     • Loss of seasonal ranges; habitat fragmentation; fencing; poaching; loss of resources

4. Species distribution models - case study: moose in Scandinavia

   • 1.1 Spatial autocorrelation
     • We want to find out moose distribution during winter
     • Count moose droppings - the more droppings, the longer they spent there; food: scots pine and field layer
   • 1.2 Results
     • Count data are clumped
     • Non-spatial analysis - 20% explained; large scale - 29% explained; small scale - 64% explained
   • 1.3 Spatial autocorrelation
     • Habitat availability over- and under-estimated presence of moose in the landscape
     • The distance to the nearest good and poor patch affects the probability of presence of moose
     • The scale of the distance depends on the species and the environment

5. Point processes and distribution - case study: saiga antelopes in Central Asia and Russia

   • 2.1 Drivers of migration
     • Resources (food, water); predation; parasites; competition; climate; people
   • 2.2 Data and methods
     • Aerial and ground surveys - spring, summer and winter; precipitation; NDVI (normalised difference vegetation index) at 16 day intervals
       • Near-infrared radiation minus visible radiation divided by near-infrared radiation plus visible radiation
   • 2.3 Results
     • Saiga antelopes follow NDVI (vegetation) south in winter; don't follow NDVI in summer to the north as this is where the people are
   • 2.4 Conclusion
     • A single proxy for resources (NDVI) explains the distribution of saiga antelopes
       • NDVI is spatially and temporally explicit; NDVI is linked to rainfall and seasonal climate
     • The effect of humans remains to be studied

6. Protected areas and movement - case study: wildebeest in the Serengeti, Tanzania

   • Migration of wildebeest in the Serengeti (1.6 million); large scale movements throughout the year; human population increase - land use change (agriculture), poaching
   • 3.1 Data and methods
     • GPS collars on wildebeest
     • Mapping of protected areas and their protection status - Serengeti National Park: tourism; conservation areas: tourism, settlement, livestock (hunting); gave reserves: tourism and hunting
   • 3.2 Results
     • Wildebeest spend considerable time outside protected areas
   • 3.3 Discussion
     • New Serengeti highway will cut through the migration route

7. From point processes to movement

   • So far: locations at specific points in times - new technology (GPS) decrease time between locations (down to 2 mins)
     • 4.1 Movement concept and scale
       • Scale of data collection depends on the question and aim of the research
     • 4.2 Too many dimensions
       • Movement data consists of three dimensions: longitude, latitude & time; from 2 location dimensions to 1; net squared displacement (NSD)
       • Movement analysis: NSD can be entered in general statistics, e.g. regression; NSD = intercept + slope * variable; variable - time, NDVI, predation risk, hunting
     • 4.3 From 2 dimensions to 1
       • Net squared displacement: distance from first point to subsequent point; routed in Brownian motion theory (physics)
       • Increases linearly with time for nomadic; asymptotic for home range
     • 4.4 Moose in Scandinavia
       • More than 300 moose collared; 76 moose for at least one year; one location per hour or more -> 5 million locations
     • 4.5 The challenge
       • Distance, timing, and duration of migration; proportion of population migrating
     • 4.7 All good but how to quantify this in biological terms
       • Advantages of non-linear mixed effects models
         • Model parameters have a biological meaning (asymptot, timing and duration of migration); fewer parameters than linear modelss thus more persimonious; provide more reliable predictions than linear models outside the parameter range
     • 4.8 Non-linear mixed effects models
       • Account for individual variation; objective and repeatable way to quantify migration; no arbitrarily chosen cut off points
     • 4.9 Results
       • Males move further than females; timing, duration and distances vary between years and individuals
     • 4.10 Conclusion
       • Yearly variation in distance, timing and duration; A unified framework for migration, no cut-off points, no splitting in time or space
         • Remaining questions: what is due to variation in the environment and what due to individual characteristics? What is the genetic basis of migration behaviour and what is learned from the mother?

8. Understanding resource use and conservation: an example from Amazonian Ecuador

   • 5.1 Why map resource use?
     • Measure sustainability of extraction; develop appropriate no-take areas or reserves; define land ownership
   • 5.2 Past approaches for mapping use
     • Interviews; accompanied trips; drawn maps
   • 5.3 Alternate approach
     • Purchase inexpensive Personal Digital Assistants with GPS capacity; design interface and load cybertracker; train community members in GPS use; can be used in conjunction with interviews
   • 5,4 Potential issues for data analysis
     • Spatio-temporal autocorrelation; central place foragers
   • 5.5 Three steps
     • Net Squared Displacement (NSD)
       • Calculate mean speed, duration and distance travelled for each type of trip
     • Biased random bridge (BRB)
       • Calculate size of each area used by each individual; calculate overlap between individuals
     • Resource utilisation function (RUF)
       • Identify habitat features associated with increased use; build predictive habitat model
   • 5.6 Study rational
     • Do hunters use the forest differently to non-hunters?
   • 5.7 Net Squared Displacement
   • 5.8Biased Random Bridge
   • 5.9 Research benefits
     • GPS allow accurate assessment of resource us areas; inexpensive
     • Benefits of analysis method: can be applied to GPS data of any collection frequency and spatial scale; can be applied when locations are missing; journeys and resource extraction can be separated; movement between locations in incorporated; all analyses can be complete in the statistical program R, which is free to download
   • 5.10 Community benefits
     • Development of skills within the community - computing, understanding of GPS, giving training; greater ownership of investigation
   • 5.11 Limitations
     • Willingness of community to participate; methods require large numbers of locations; NSD needs all parts of the curve to have enough data; analyses can be computationally time consuming

9. General conclusion

   • Spatial data are challenging
   • New technologies - GPS; remote sensing (NDVI)
   • New methods - reduce dimensions; allow uniform, scale-independent analyses