Sunday, March 9, 2008

Want to Build a Honeybee Robot? Well, How Many Skills Are Packed Into That Tiny Critter?

*eat as larvae
*grow as a larva
*defecate once, then
*spin cocoon
*pupate into an adult
*chew way out of cell!
*breath
*eat pollen from cells
*beg for and accept brood food from nurse bees
*produce saliva in salivary glands to digest food
*cleaning self, brushing debris off body
*grooming antenna
*grooming other workers
*walk
*decide what to eat for each developmental stage
*cap brood
*tend brood
*feeding larvae
*recognize full sisters vs half sisters (different father)? and preferentially feed, especially if queen larvae
*produce larva food in mandibular glands when young
*produce protein and vitamin food for larva in hypopharyngeal glands when young
*produce alarm pheromone in mandibular gland when older
*decide what food to feed larvae depending on larval age!
*recognize adult begging and feed 'em
*regurgitate food for adults
*groom queen
*feed queen
*licking pheromones off queen
*exchanging with other workers
*trophallaxis which is NOT involved in feeding but in information exchange
*antennate
*receive nectar
*produce invertase for honey production in hypopharyngeal glands
*evaporating nectar on mouth
*placing nectar in cells
*fanning nectar into honey
*produce wax
*preparing it with saliva and warming it
*cap honey cells
*packing pollen in cells
*receiving water
*build honeycomb
*occasionally move eggs/young larvae to queen cells!
*tear down queen cells when done
*recognize queen larval cells and feed royal jelly
*recognize diploid drone larvae and kill 'em
*clean debris from hive
*removing dead larvae and pupae from cells
*recognizing dead workers
*clean cells
*smooth edges
*remove caps
*sense humidity
*sense carbon dioxide levels
*ventilate hive mostly at entrance
*sense temp and maintain brood temp to close optimum between 30 and 35degC often to within .5deg
*upping metabolic rate to warm
*lining up in fanning chains to cool brood
*spreading water for cooling
*tongue lashing of water for faster cooling
*leaving nest to reduce metabolic heat and leave more room for ventilating
*patrolling (collecting and transferring info)
*resting
*guard duty
*release scent pheromone from nasanov gland at entrance
*letting in bees of same colony
*letting in workers of foreign colony if proper submissive behaviors especially if they bring food
*(foreign bees may be adopted, what behaviors? learning new hive scent!)
*recognizing adversaries
*releasing alarm pheromones if alarmed by attacker
*responding to alarm pheromones and flying out searching for attacker
*attack by buzzing
*attack by biting
*attack by pulling hairs
*group grappling on hard shelled insects
*turning and kicking and fanning to attack ants
*shimmering to attack wasps
*attacking robber honeybees at entrance in midair flight
*attack by stinging
*pulling away from stinger
*stinger continues to pump poison for a minute
*stinger exudes alarm pheromones
*flying to attacker based on sting pheromones
*zeroing in on sting pheromone and stinging in close spot
*recognizing nest mates
*orientation flight 5 minutes of circles to familiarize with territory
*first voiding of feces!
*fly!
*find new source of resource, and come back
make round dance for sources closer than 15m communicating source quality compared to colony needs
*observe round dances
*make waggle dances for sources farther than 100m (intermediate dances for intermediate sources!)
*communicate amount of flight energy expended (distance and headwind), food quality vs needs, direction
*observing waggle dances
*summing info from multiple dances from same source!
*comparing info from dances from many sources and deciding on the most efficient
*squeak during observation of dance for food samples
*stock up on honey for energy before a flight take more if waggle dance says long distance
*DVAV dance to call workers to the dance stage
*there are probably other movement oriented communication acts
*keeping track of how many bees have taken pollen/nectar/water from you
*deciding to switch resources to forage for
*deciding whether to look for a dance to follow or go on a new scouting run
*(I don't understand, all the options a forager has in her 4 -5 days of foraging, when do you scout, when follow a dance, when go to a site you've memorized? age dependant?)
*more experienced foragers preferentially scout
*decide to scout when colony needs are NOT being met
*up to 35% of total foragers may scout!
*don't spend TOO much time on any foraging trip so that you can maximize time in nest *gaining info about changing foraging needs and possibilities
*deciding whether to unload pollen from incoming foragers
*deciding whether to unload nectar from incoming foragers
*deciding whether to unload water from incoming foragers
*deciding whether to unload propolis from incoming foragers
*sensing compass angle of sun
*sensing compass angle of polarization
*using landmarks to navigate
*sensing magnetic field
*assessing hierarchy of information between these
*compensating for passage of time
*remembering location of source over night
*learn the times of day for multiple optimal nectar flows
*visit specific learned patches at specific times of optimal nectar flows each day
*decide to do this instead of responding to other input
*visit same small patch at a foraging site repeatedly
*sense wind speed
*recognizing flowers visually
*color vision
*pattern vision
*pattern memory
*learn most efficient way to approach and collect food from a flower in 1 to 5 visits
*stick to that method.
*workers can memorize these things over 175 days of winter confinement!
*flower preference and food type preference is partially hereditary and can be bred for!
*visit mostly a single species of flower per foraging trip
*visit that species on next trip
*from 1 to >1000 flowers per trip
*when yield is good make short hops and change direction often between flowers to remain in one patch
*when collection efficiency drops decide to make longer hops not changing direction to move out of patch
*10 to 150 trips per day
*on average travel further for pollen than nectar (lighter?)
*sensing movement: flicker fusion frequency of 300 stripes/sec (human is 15-20)
*recognizing flowers olfactorily
*sensing olfactory direction
*foraging (lasts only about 800km 4 -5 days 10 trips /day)
*up to 6miles distant! (10km)
*recognizing pollen
*choose pollen based on odor and configuration of pollen grains
*collecting pollen
*manipulate anthers to get grains
*insert proboscis into tubular flowers for nectar, getting pollen on head and legs
*force closed flowers open and find pollen inside
*run along spikes and catkins getting pollen on body
*press abdomen against presentation flowers to get pollen
*packing it on pollen basket making it sticky with honey
*taking pollen packets off and placing in cells
*treating pollen with phytoacids and enzymes
*collecting nectar
*work flowers from the side for nectar avoiding pollen
*find side holes bitten by bumblebees for nectar
*learn to avoid getting trapped by flowers like alfalfa and search for 'tripped' flowers
*or approach trap flowers from the side
*collecting water
*releasing nasonov scent at resource location
*deciding to release nasonov scent at resources based on colony needs and source quality
*recognizing foreign nests as honey sources at end of summer or in time of dearth
*pre attack flights at nest entrance
*gaining entrance, ingesting honey and bringing back to colony
*(receiving bees observing waggle dance can tell by honey that it's from a hive?)
*successful robbers more likely to seek out more hives for honey
*regurgitating nectar and water
*using proventriculus to select contents of crop to digest
*excrete while outside nest
*hold it in in the winter!
*collecting plant resins
*produce propolis
*coat nest cavity with propolis
*plug holes with propolis
*embalm carcasses of mice etc.. with propolis
*develop into a queen
*as queen, produce queen substance in mandibular gland when a queen
*as queen, lay eggs
*as queen, sense difference between worker and drone cells
*as queen, fertilize an egg or not depending on which kind of cell
*as queen, recognize swarming behavior and leave hive with workers!
*develop into drone
*as newly hatched or prehatched queen, release pheromone and pipe
workers listen to piping
*don't let any other virgin queens out if hear a new virgin piping in the nest
*as queen, if more than one virgin queen in nest kill rival queens or kill queens still in cells
*workers form ball around loser in combat and finish her off.
*pre mating flight workers prepare queen
*workers push queen out of nest
*workers form dense cluster at entrance fanning nasonov pheromones to call her back
*as queen, go on mating flights
*as queen, find mating area FAR from nest probably >2km
*as drone find mating area FAR from nest probably >2km
*drone recognize queen
*queen recognize drone
*try to go on flights when weather conditions are right
*as drone follow queens
*as drone mount queen
*as queen decide whether to be receptive (within seconds?)
*as drone mate with queen and die
*as second drone push mating plug aside to mate with already mated queen
*when queen returns workers lick mating plug and remove it
*if not enough sperm go on more flights
*after enough sperm squeeze most of it out retaining enough in spermatheca.
*workers eat excess sperm off queen.
*when resources are low expel drones from nest
*if resrources are high or it's autumn and no queen keep drones for spring
*if queen is injured, initiate supersedure
*if queen is dies or drops in pheromone production, initiate supersedure
*if queen lays insufficient number of eggs, or lays unfertilized eggs, initiate supersedure
supersedure is to raise half a dozen new queens
*in cases of old queen laying fertilized eggs but in some way inferior, let new queen go on mating flight, keep old queen alive UNTIL new queen proves herself to be able to lay fertilized eggs, THEN let new queen kill old one.
*Or let old queen attempt swarm? will workers go with her? and practice new supersedure in new colony? not likely.
*in absence of queen and brood workers will begin to lay eggs
*laying workers may fight each other
*colonies with laying workers may raise up drones
*sense worker and larvae population levels
*begin brood rearing in dead of winter (signal?)
*synthesize information from cues about: colony size, brood area, adult population, brood nest congestion, worker age distribution, reduced transmission through nest of queen pheromones to begin queen rearing and preswarm conditions
*build drone cells numbered in proportion to colony resources
*feed queen more before swarming to induce more egg laying
*but closer to swarming time, feed her less, so she becomes light enough to fly
*harass queen prior to swarming
*younger workers engorge with honey during preswarming conditions up to 10 days prior, to be prepared for quick swarming decision
*sense worker brood maximum (all cells full) to initiate preswarm queen production
*initiate queen cells for swarming
*destroy queen cells if conditions aren't right
*sense weather conditions to decide on swarming
*engage in positive feedback loop swarm initiation
*harass queen during swarm initiation
*form swarm around queen
*fly with the swarm if you are young (less than 10days old), not old. (what about the scouts?)
*if on flight but queen didn't come, go back!
*form swarm cluster at interim site hanging from branch etc..
*scout for nests (initiated prior to swarm) (how to decide who scouts?)
*scouts study inside of possible nest
*scouts study environs of nest
*a scout synthesizes info about nest site from different times of day?
*scouts waggle dance site position and quality
*observing nest selection dances
*choose one and fly to it becoming a scout
*when consensus is reached (threshold level?) scouts perform buzzing run and initiates swarm flight
*when nest site is decided on form large cloud of bees and follow scouts
*scouts perform streaking flights through the cloud to guide it
*scouts attract workers to site entrance with nasonov glands
*founding nest
*initiate comb building
*huddling and buzzing to warm the wax to optimal working temp
*sense gravitational gradient
*if population levels warrant it go in an afterswarm with a newly mated virgin queen. test whether she lays fertile eggs first?
*covering nest entrance in winter!
*clustering in nest begins with ambient temps of 18degC
*at 14degC cluster develops compact shell of a few bees thickness the bees are in dynamic equilibrium circulating between layers
*by -5degC the cluster can no longer contract
*bees break out of cluster to feed on honey
*maintain minimal inner cluster temp of 13degC outer temp 8degC, maximal temp of 20degC

259 skills

-mostly from Winston's "Biology of the Honey Bee"

How is it ALL packed in there?


3 comments:

Terra Maple Forester said...

I know people who program robots. They are still working on how to get a robot to pick things up. Nature is way ahead of humans. So many people take for granted what nature can do, but you take the time to really see it. To me, that's pantheism, that wonder in the face of nature.

barry goldman said...

pantheism... at least vs theism... the pan part vs the theos part is to ask whether nature was created by and is directed by a single monolithic mind like process.

(not that we even have a clue how OUR minds work, and for sure they are NOT as monolithic as we think. and the process of wondering at nature that is science is beginning to dig in and untangle all those neurons and try to untangle subrpocesses apart and...)

so we start to LOOK at nature around us and we see parts. MANY parts. and then we start to tease apart the INTERACTIONS between parts.

and then from the other end we learn how to build up complex processes from the interactions between simple parts, as we do in digital electronics and building robots and programming computers.

as we go up and down the scale, from the top down, watching a honeybee, and from the bottom up, TRYING to build honeybee robots, we get closer to the meeting point, realizing that nature is VERY pan, lots of interacting agents, and not a single monolithic mind.

it is the goal of the complexity lab manual, to give the experience of working up and down this scale of interacting parts.

Terra Maple Forester said...

I have not really thought of pantheism in that way. In my experience, pantheists think of nature or the universe as God, meaning 1) the thing that makes us feel wonder and awe, and 2)life being created by evolution, things happening as they do because of nature and not because of the control of a sentient mind. But many pantheists tend to think of the universe or nature or evolution in a more monolithic way. (There's a description of pantheism at http://www.pantheism.net/manifest.htm.) I like the way you convey it as many different things. That makes it all the more miraculous.

At first, it makes me think of the power of cooperation. Like a dance, a whole bunch of people doing different things to crate a larger whole. It seems like a beautiful thing, the greatness we can create if we all work together.

But then I think we got here with evolution by pursuing our own interests, by propogating our own genes, not by trying to trying to build things together.

We animals and plants pursue our own ends, and the interaction of all these things pursuing their own ends creates something that none of us individually planned or could conceptualize.