Sea otter diets population status Causes and consequences of individual foraging specializations M Tim Tinker USGS BRD Western Ecological Research Center and EE Biology UC Santa Cruz Consumer prey interactions represent one of the central themes of ecology Today s Presentation PART 1 Individual specialization as a behavioral response to reduced food availability Primary Collaborators J Estes and G Bentall PART 2 Conservation implications of individual foraging specializations a case study Primary Collaborator C Kreuder Johnson Niche breadth and the economics of consumer diet choice In general predators are expected to select prey in such a way as to maximize their lifetime fitness Theoretical and Empirical studies suggest that dietary diversity predator niche breadth increases as food availability decreases Optimal Diet Choice Rank potential prey according to rate of energy gain e1 e3 e2 h1 h2 h3 Add prey types to diet until j e i i i 1 j 1 i ei e j 1 h j 1 i 1 Prediction when profitable prey types are abundant predator diet breadth should be narrow Expected Diet Population level Low Predator Density High Predator Density Empirical support from cross sectional dietary data fecal composition and stomach content analysis etc Assumption diets of con specifics are similar Alternative Explanation Low Predator Density High Predator Density In this scenario same population pattern produced by specialization and diversification of individual diets How to distinguish between these two scenarios Fairly Stringent Requirements experimental manipulation of food abundance control for environmental effects age sex effects genetic differences longitudinal dietary data from individuals Few opportunities to test particularly for large vertebrates Foraging Studies of Sea otters along Central Coast of California Easy to observe feeding o Near shore distribution o Bring all prey to surface Resource limited pop n o o o o High mortality rates Poor body condition High time feeding Very diverse diet A Fortuitous Experiment Central CA range 1982 Recovery Plan recommended establishment of a secondary colony San Nicolas Island was selected as preferred site Between 1987 1990 sea otters were translocated from central CA to San Nicolas Island N W E S San Nicolas Island 200 0 200 400 Kilometer San Nicolas Island Remote population distinct from the mainland range RC Population still small N 35 animals food abundant Current sea otters SN only 1 2 generations removed from founders Perfect candidate for comparative studies low population density at San Nicolas SN vs high population density in range center RC Sea Otter Density km2 3 0 2 5 2 0 1 5 1 0 0 5 0 0 Hypotheses At low population density when food is abundant individuals converge on single best diet small niche breadth At high population densities when food is limiting there is increased niche breadth H1 greater niche breadth caused by increased within individual diet diversity diet generalists H2 greater niche breadth caused by increased among individual diet diversity diet specialists Study Methods Sea otters were captured radiotagged then monitored by telemetry and direct observation Study Methods Sea otters were captured radio tagged then monitored by telemetry and direct observation Study Methods Sea otters were captured radio tagged then monitored by telemetry and direct observation Range Center RC 72 animals detailed diet data from 34 San Nicolas I SN 23 animals detailed diet data from 11 Checking our assumptions Q Are food resources truly more abundant at San Nicolas Island If so we might expect benthic density of key prey Higher rate of energy gain Less time devoted to feeding Better individual body condition chunky otters Energy Rate kcal min 1 0 1 0 001 20 15 10 5 0 Proportion of Activity Feeding 0 01 RC SN Mass kg Length cm 2 Urchin Density per m Checking our assumptions 10 0 5 0 4 0 3 0 2 0 1 0 0001 0 0 25 0 25 0 20 0 15 0 10 Comparison Results Total Niche width at SN was 40 smaller S Windex 1 16 at SN S Windex 1 98 at RC At SN 90 of niche breadth represented within individual diet diversity while at RC there was 3x more among individual diversity At SN each individual s diet resembled the population average PSI 0 82 At RC individual diets varied from population average PSI 0 57 PSI PSI Proportional ProportionalSimilarity SimilarityIndex IndexThe Thedegree degreeto to which whichindividual individualdiets dietsresemble resemblethe theaverage averagepopulation population diet diet with withaavalue valueof of11indicating indicatingcomplete completeoverlap overlap San Nicolas Island SN 0 8 Range Center RC Population average diet low diversity Population average diet high diversity 0 5 0 7 0 4 0 6 0 5 0 3 0 4 0 2 0 3 0 2 0 1 0 1 0 0 ur 0 8 kc ga cc Individual 1 SN 0 6 cl lo ab sk cm st kc 0 6 cc cl ga sk mu ur st wo ab sd cm Individual 1 RC 0 4 0 4 0 2 0 2 0 0 8 0 Individual 2 SN 0 6 0 6 Individual 2 RC 0 4 0 4 0 2 0 2 0 0 8 0 6 0 Individual 3 SN 0 6 Individual 3 RC 0 4 0 4 0 2 0 0 2 0 H2 increased niche width at high population density caused by increased among individual diet diversity specialization Why would animals specialize at low food availability Profitable preferred prey is easy to handle less profitable prey requires specific skills for efficient utilization Learning inertia individuals learn prey specific skills which are not easily transferred to alternate prey Result different prey types associated with distinct foraging strategies behavioral polymorphism Predictions for behaviorallymediated foraging polymorphism 1 Increased Diet Specialization at RC should be associated with increased variation in foraging behavior 2 Specialists should enjoy performance advantage for their preferred prey Behavioral Variables Dive depth duration bottom time Duration of post dive surface interval Dive ascent descent rate Variation in dive parameters Frequency of successful dives capture success Edible biomass prey item Number of items dive Handling time per item Frequency of Tool use Other data prey stealing etc Cluster Analysis of Behavior Variables On the basis of diving and foraging behavior otters could be grouped into 4 clusters Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case Case 2 18 36 7 19 27 31 3 21 14 25 26 30 22 15 24 20 28 16 5 10 1 12 13 6 17 32 29 8 11 42 34 9 4 33 23