fuchsbau

Ant Encounters, Deborah Gordon

(back to books)

---

• Substrat: Ant colony behaviour arises from dynamical networks of interaction.
  • use local information; don't tell each other what to do; nobody understands needs on macro level
  • living systems cause own development
  • information = event that produces change of state
  • ants not individuals in ecological sense (individual ant doesn't reproduce); instead, colonies make more colonies; ant population is all nearby colonies
  • abuse from others: spiders mimic cuticular hydrocarbon profile to enter nest & eat larvae; spore-producing fungus infects/kills ants; queens inflitrate/take over nest
• in complex adaptive systems local interactions determine global outcomes
  • patterns of interactions more important than content
  • enough ants must perform well often enough (perfection not necessary)
  • repeated behaviour is quasi-long-term memory (e.g. colonies "remember" trails for decades)
  • ants with outsized interaction rate have outsized impact on colony behaviour
  • ants regulate interaction rate (e.g. large colonies have idle ants which might buffer)
  • intuitive rule for ant behaviour: "if I meet ant with odor A ~3 times in next ~30 sec., probability I forage +10%, else -20%."
  • smaller groups use straighter paths (algorithm: when ant meets another, more likely to turn randomly)
• colony growth
  • foundation: mating flight (reproductives from nearby colonies meet to mate), mated queen starts nest, lays eggs; initially, queen feeds/cares for brood
  • eggs grow into larvae and then pupae (inside the nest); ant emerges from pupal case & doesn't grow further; colony directs most food to feed larvae
  • colony growth has sigmoid shape (initial growth feeds growth because more workers)
  • high initial death rate; harvester ants: if 10 k ants & 2y colony, likely survives >20 ys
  • foragers don't scale linearly (harvester ants: 50% (20%) in 1 k (10 k) colony); instead, more idle ants over time (excess capacity; maybe buffer for interactions)
  • colony mature when it begins to produce winged reproductives
  • older colonies use patrollers to avoid other colonies; younger colonies return to foraging place even when there is colony overlap (maybe due to growth curve)
• core ant processes: 1) leave nest, find food, bring back; 2) build/repair nest; 3) feed/groom larvae, move pupae.
  • other processes (examples): grow/harvest fungus; protect/harvest aphids; move nest; raids; alarm; recruit; found new colony
  • if nothing drastic happens, ant does what it did previous day
  • patrolling: stimulated to leave nest by morning sunlight/higher temperatures; put down chemical line (~20 cm) on nest mound to guide foraging
  • foraging: stimulated to leave by interactions with returning patrollers or returning foragers; foragers drop food at nest entrance & wait
    • foragers don't switch back (sink); young ant is source, starts task inside (close to birth place), may end up at entrance & become forager
    • ~10% of foragers ("star foragers") make many trips, majority only few; when removing star foragers, other foragers become star foragers
  • colony move: ants investigate new sites, more likely to stay in viable sites; if density high, scouts recruit rest of colony to move
  • alarm: ants disperse alarm pheromone in air; alarmed ants run in circles, sending out more pheromones, getting more ants to run around.
• general ant info
  • older than dinosaurs (~140 mio. ys old); ancestors are wasps; ants = 33% of mass of all insects (2% of insect species)
  • olfaction very important (perceive odors from objects touched with antennae); ~15 glands, each secretes different chemicals; sensitive to vibration
  • ants can be mutualists: e.g. ant-plant-aphid relationships
  • many variations to basic lifecycle: slave-making; daughter replaces queen; queens live together; several queens build nest, fight to death; ...