British Ecological Society 2016

Presentation on the influence of environmental regime on resource selection at the British Ecological Society Conference, Liverpool, December 2016. Here, we show that environmental predictability influences whether seabird foraging behaviour is influenced by static or dynamic features.

Transcript

1. Does environmental predictability play a role in shaping individual foraging behaviour? Alice M Trevail, Jonathan A Green, Jonathan Sharples, Jeffrey Polton, Peter Miller, John Arnould & Samantha C Patrick @AliceTrevail alice.trevail@Liverpool.ac.uk
2. Niche specialisation Generates & maintains biodiversity Facilitates speciation Promotes coexistence Driven by: • Competition • Ecological opportunity Niche specialisation @AliceTrevail Matt Doggett
3. Individual specialisation Individual specialisation in habitat selection • Underpin niche specialisation • Varies within & between species & populations ` Ed Schneider The Telegraph Liam Quinn Anders Lundberg Antarctic connection aquaportail.com Antarctica.gov.au Joachim S Muller Robert OToole @AliceTrevail
4. Individual specialisation Extrinsic e.g. Resource availability ` Ed Schneider The Telegraph Liam Quinn Anders Lundberg Antarctic connection aquaportail.com Antarctica.gov.au Joachim S Muller Robert OToole Intrinsic e.g. hereditary Individual specialisation in habitat selection @AliceTrevail • Underpin niche specialisation • Varies within & between species & populations
5. Resources = source of competition & opportunity Specialisations = arise through consistent behaviours Q : Do predictable resources lead to specialisations? Resource availability Richard Shucksmith @AliceTrevail
6. Physical parameters = Proxy for resources e.g. Depth & currents Upwelling regions Resource hotspots Q : Does a predictable environment lead to specialisation? Environmental predictability Satellite image: NOC Phytoplankton bloom @AliceTrevail
7. Environmental variables Latitude (°N) Longitude (°W) Depth (m) 54 53 52 6 5 4 3 250 200 150 100 50 0 Static 1. Depth Currents Water column structure @AliceTrevail
8. Static 1. Depth 2. Tidal Stratification Nutrient availability Productivity Environmental variables Stratified water column = more productive Mixed water column = less productive @AliceTrevail
9. Environmental variables Static 1. Depth 2. Tidal Stratification Dynamic 1. Fronts Upwelling Increase productivity @AliceTrevail Front ` Kent Smith
10. Predictability at study sites Skomer Puffin Island Skomer Puffin Is. Vertically stratified, Stronger fronts & Spatially heterogeneous Predictable resource hotspots Vertically mixed, Weaker fronts & Spatially homogenous Unpredictable resource distribution @AliceTrevail
11. Methods: seabird tracking • Seabirds = Model species for behaviour studies Colonial high competition niche segregation • Study species = Black legged kittiwake • GPS loggers (iGotU 120) • Multiple trips per individual • Puffin Island: 505 trips, 59 individuals, 6 years • Skomer Island: 22 trips, 10 individuals, 1 year Latitude (°) 4.4 4.2 4 53.3 53.4 53.5 53.6 4.3 4.1 Puffin Island 52 51.9 51.8 51.7 Skomer Island 5.5 5.4 5.3 5.2 Longitude (°) @AliceTrevail
12. Methods: quantifying resource selection 5.6 5.4 5.2 51.6 51.7 51.8 51.9 52 52.1 Latitude (°N) Tidal stratification Longitude (°W) 7 6 5 4 3 2 1 Tidal stratification log10 (m-2 s3) @AliceTrevail
13. Methods: quantifying resource selection 5.6 5.4 5.2 51.6 51.7 51.8 51.9 52 52.1 Latitude (°N) Tidal stratification 5.6 5.4 5.2 Stratification for GPS points = Used (1) Longitude (°W) 7 6 5 4 3 2 1 Tidal stratification log10 (m-2 s3) @AliceTrevail
14. 5.6 5.4 5.2 Methods: quantifying resource selection 5.6 5.4 5.2 51.6 51.7 51.8 51.9 52 52.1 Latitude (°N) Tidal stratification Stratification for random points = Available (0) 5.6 5.4 5.2 Stratification for GPS points = Used (1) • Binomial GLMM with used/available (1/0) as response variable Longitude (°W) 7 6 5 4 3 2 1 Tidal stratification log10 (m-2 s3) @AliceTrevail
15. 0 20 40 60 80 100 120 0.00 0.02 0.04 Resource selection: Static (depth) Predictable = Skomer Island Available Density Unpredictable = Puffin Island Available 0.00 0.01 0.02 Depth (m) @AliceTrevail
16. 0 20 40 60 80 100 120 0.00 0.02 0.04 Resource selection: Static (depth) Predictable = Skomer Island Available Used Density Unpredictable = Puffin Island Available Used 0.00 0.01 0.02 Depth (m) @AliceTrevail
17. Resource selection: Static (depth) 0 20 40 60 80 100 120 0.0 0.2 0.4 0.6 0.8 1.0 Depth (m) Probability of habitat use @AliceTrevail Predictable (Skomer Island) = slope significant (p < 0.05) Unpredictable (Puffin Island) = slope significant (p < 0.05) Stronger selection for this static feature at the unpredictable environment
18. Resource selection: Static (stratification) 0.0 0.4 0.8 1.2 Density 0 1 2 3 4 5 6 0.0 0.5 1.0 1.5 2.0 Tidal stratification (log10 (m-2 s3)) @AliceTrevail Predictable = Skomer Island Available Unpredictable = Puffin Island Available
19. Resource selection: Static (stratification) 0.0 0.4 0.8 1.2 Density 0 1 2 3 4 5 6 0.0 0.5 1.0 1.5 2.0 Tidal stratification (log10 (m-2 s3)) @AliceTrevail Predictable = Skomer Island Available Unpredictable = Puffin Island Available
20. Resource selection: Static (stratification) 1 2 3 4 5 6 7 0.0 0.2 0.4 0.6 0.8 1.0 Tidal stratification (log10 (m-2 s3)) Probability of habitat use @AliceTrevail Predictable (Skomer Island) = slope NOT significant (p = 0.05) Unpredictable (Puffin Island) = slope significant (p < 0.05) Stronger selection for this static feature at the unpredictable environment
21. Resource selection: Dynamic (fronts) Cross-front gradient strength (°C/1.2km) 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0 20 40 60 0 100 200 Density @AliceTrevail Predictable = Skomer Island Available Unpredictable = Puffin Island Available
22. Resource selection: Dynamic (fronts) Cross-front gradient strength (°C/1.2km) 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0 20 40 60 0 100 200 Density @AliceTrevail Predictable = Skomer Island Available Unpredictable = Puffin Island Available
23. Resource selection: Dynamic (fronts) @AliceTrevail 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.0 0.2 0.4 0.6 0.8 1.0 Cross-front gradient strength (°C/1.2km) Probability of habitat use Predictable (Skomer Island) = slope significant (p < 0.05) Unpredictable (Puffin Island) = slope significant (p < 0.05) Stronger selection for this dynamic feature at the predictable environment
24. Predictability drives resource selection Shallower Deeper Mixed Stratified Predictable = Skomer Island Unpredictable = Puffin Island Weaker fronts Stronger fronts Static Dynamic @AliceTrevail
25. Predictability drives resource selection @AliceTrevail Driven by opportunity to specialise? Heterogeneous environment Homogeneous environment Predictable hotspots Unpredictable resources Specialised on dynamic features Specialised on static features
26. Predictability drives resource selection @AliceTrevail Driven by opportunity to specialise? Heterogeneous environment Homogeneous environment Predictable hotspots Unpredictable resources Specialised on dynamic features Specialised on static features Consequences under environmental change: If adapted to static features = less responsive to change?
27. Selection of weaker fronts Both colonies select for weaker fronts > Transient nature of ocean fronts > Short duration = renewed supply of resources Future work: explore alternative front metrics, including persistence @AliceTrevail
28. Next step: Individual resource selection Individual specialisations underpin niche diversification Evidence from models suggests individual variation in resource selection pressures Q: Does environmental predictability influence strength of individual specialisations? @AliceTrevail
29. Thanks to Puffin Island field teams, including: Federico de Pascalis, Harriet Clarke, Ruth Dunn, Phil Collins & Louise Soanes As well as Skomer wardens, B and Ed, & Ros Green
30. Thank you! Questions? @AliceTrevail alice.trevail@Liverpool.ac.uk
31. Predictability at study sites Skomer Puffin Island Skomer Puffin Is. Depth 52 51.8 51.6 5 5.2 5.4 5.6 0 160 Depth (m) 5 4 4.5 53.6 53.4 53.2 Depth Deeper & more variable Shallower & spatially homogenous Latitude (°N) Longitude (°W) Longitude (°W) @AliceTrevail
32. 0 8 Tidal stratification log10 (m-2 s3) 7 6 5 4 3 2 1 0 8 52 51.8 51.6 5 5.2 5.4 5.6 Tidal stratification Tidal stratification log10 (m-2 s3) 7 6 5 4 3 2 1 0 8 5 4 4.5 53.6 53.4 53.2 Tidal stratification Tidal stratification log10 (m-2 s3) Predictability at study sites Skomer Puffin Island Skomer Puffin Is. Higher stratification index & more variable Vertically mixed & spatially homogenous Latitude (°N) Longitude (°W) Longitude (°W) More mixed More stratified @AliceTrevail
33. Predictability at study sites Skomer Puffin Island Skomer Puffin Is. @AliceTrevail Stronger fronts & spatially more variable Weaker fronts & spatially homogenous 52 51.8 51.6 5 5.2 5.4 5.6 Frontal strength 5 4 4.5 53.6 53.4 53.2 Frontal strength Latitude (°N) Longitude (°W) Longitude (°W) 0 0.12 Cross-front gradient strength (°C/1.2km)