Behavioural Ecology:
Revision Notes

Compiled as a third-year student at the University of Edinburgh, based on the information given in lectures.
General points are shown in normal font; specific examples are given in italics.

• Some complex behaviours are genetically based. Migratory restlessness in warblers: high in Germany, low in Canary Islands where they don't migrate, intermediate in hybrids (h²=0.45). Migratory response can be altered by artificial selection.

Sexual selection

• May be intra-sexual (fighting for mates) or inter-sexual (showing off to mates).
• Elephant seals: males are large and have bulbous noses. Females can't stop a male mating with them but can 'protest' (which is more likely with unattractive males), encouraging other males to intervene.
• Dunnocks: different mating systems: monogamy, polyandry, polygyny, polygynandry. Chicks survive best in polyandrous families where alpha and beta males both help feed chicks. % of feeds given by beta male depends on his access to female (and chances of paternity). Females try and surreptitiously mate with beta male to ensure his co-operation. Alpha male guards female to prevent this, and pecks her cloaca before mating to force ejection of past sperm. Males' and females' reproductive success depends on mating system - their interests don't coincide.
• Lions: a coalition of brothers may take over a pride. It is advantageous for females to rear young together, producing coalitions of brothers. Males taking over a pride commit infanticide, which brings females into oestrus again.

Lekking

• Observed in 7 mammal and 35 bird species (e.g. fallow deer, grouse, birds of paradise, humans).
• Success is skewed towards dominant males, though inferior ones may get 'sneaky' matings.
• Reasons for lekking: males congregate at 'hotspots' where females are likely to be? reduces risk of predation? 'hotshot' males attract others? testing ground for good genes?

Inter-sexual selection

• Males may offer females resources (e.g. fly given as food by scorpion fly) or offer only genes.
• Long-tailed widow birds: longer tails give mating success; can be demonstrated by manipulation.
• Sedge warblers: larger song repertoires are attractive in males; can be demonstrated with taped songs.
• Fisher's runaway selection: equilibrium is unstable.
• Covariance between male and female traits (in sticklebacks, son's intensity of red colouration is related to daughter's preference for red males).
• Mate preference can be genetically determined (e.g. red/melanic preference in 2-spot ladybirds).
• Stalk-eyed flies: long eye stalks attract females; artificial selection can shorten but not lengthen stalks - has natural selection 'used up' the available variation?
• Sexy son hypothesis - do females behave like this because they want attractive sons?
• Not in pied flycatchers: repeatability of mating success is low (maximum h² = 0.2) and polygyny is rare.
• Handicap principle - males prove their worth with characteristics that hamper their survival.
• Females benefit by choosing males with 'marker' if marker doesn't reduce male's fitness below average fitness of unmarked individuals. Selection can produce larger markers if fitter individuals invest only part of their superiority in a larger handicap.
• 'Showiness' in birds may be related to susceptibility to blood parasites
• Brighter male bluebirds breed earlier, care better for young, and produce more offspring.
• Swallows with longer tails return from Africa earlier, breed earlier, have more second broods, have fewer 'fault bars' (signs of malnutrition) in feathers or broken feathers. Experiments show that long tails hamper food-catching ability. Parasite load affects tail length next season (demonstrated by moving mites between nests). Mite resistance is genetic (demonstrated by moving chicks between nests).

Mixed reproductive strategies

• Field crickets: loud singing attracts both mates and parasites; 'satellite' males are silent but get mates by hanging around loud males and intercepting visiting females.
• 'Transvestite' scorpion flies steal flies from males by pretending to be females.
• Coho salmon: 3 year-old males develop big hook nose to attract females, but 2 year-old males can remain as tiny 'jacks' forever - two possible life-history strategies. Jacks more likely to survive to breeding age, but have shorter breeding period and can't get so close to females. Overall success rate of jack and hooknose strategies are similar, which must be the case for the two strategies to persist.

Altruism

• Tasmanian native hen: females may share a territory with two males. If males are brothers, they should co-operate if beta's gain in young is more than twice alpha's gain. If males are unrelated, alpha should evict beta.
• Inclusive fitness (IF) = 1 + rb - c (where b is benefit, c is cost, and r is relatedness).
• Altruism will spread if 1 + rb - c > 1 (in case of full siblings, where c < ½b).
• Altruism among unrelated individuals is not an Evolutionarily Stable Strategy.
• Siblings are selected to be altruistic when c < ½b and selfish when c < 2b.
• Parent-offspring conflict - parents want siblings to co-operate when c < b.
• Weaning conflict - parent wants to stop young suckling before young want to stop - prolonged suckling reduces survival of future siblings.
• Reduction in mortality necessary to select for brood reduction in nests is highest for siblicide, lower for infanticide, and lowest for suicide.

Reciprocal altruism

• Vampire bats: those that fed well may regurgitate blood to feed those that didn't. Most common among related individuals, but also occurs among unrelated ones that associate regularly with one another (high 'association index'). Bats are long-lived and have long-memories - they remember who has helped them. Benefit to a near-starving bat is greater than the cost to a well-fed one.

Kin selection and co-operative breeding

• White fronted bee-eaters: breed in 'teams' of related individuals.
• Grey-crowned babblers: experimentally removing helpers from group reduces number of young fledged.
• Parents themselves make a lower contribution to feeding effort in larger groups, helping parents' long-term survival and fecundity.
• Probability of helping depends on relatedness, but even unrelated individuals may help if this help is reciprocated later.
• Next year's siblings will be full siblings (and more inclined to help) only if both parents survive, so helping is most common in long-lived species.
• Florida scrub jays: genetic gain of helper is 0.14, first-time breeder is 0.62. So why help?
• Other possible benefits of being a helper: gain experience, inherit parental territory, gain a mate (unrelated helpers may usurp breeders), group dispersal (as a coalition), reciprocal help.
• Pied kingfishers: on Lave Naivisha, there is typically 1 helper, son of breeding pair. On Lake Victoria, where conditions are harsher, there is a 2nd helper who is unrelated and later takes over breeding female.
• Green wood hoopoes: helpers form coalitions with younger siblings for group dispersal later.
• Breeder will want help, but helper will usually want to breed alone if it has a good chance.
• Acorn woodpeckers: store acorns in holes in tree trunks, requiring huge maintenance effort. Higher proportion of young are retained as helpers when territory is scarce.
• Splendid fairy wrens: shortage of females encourages birds to become helpers.
• Index of kin selection (I_k) = increase in young produced as a result of helping
• Why do sons, not daughters, stay and help (in birds)? One sex must disperse to avoid inbreeding. Males know they are related to siblings but may be unsure about their offspring.

Communal breeding

• Multiple pairs breeding in same nest.
• Groove-billed ani: dominant female breeds last, having tossed out rivals' eggs.
• Benefits of communal breeding outweigh risks, but it is worthwhile for dominant birds to 'pay' by allowing some of rivals' chicks to survive.

Seasonally-varying sex ratios

• In bivoltine species (2 generations per year), sons are more valuable in first season because they can mate with females from second season. In second season, daughters are preferred.

Eusociality

• Several adults together in a group; overlapping generations; co-operation in nest-building, defence and brood care; reproductive dominance, sterility or distinct castes.
• Sterile castes may work for their mothers (eusociality) or for their sisters (semi-sociality).

Eusociality in Hymenoptera (bees, ants and wasps)

• Eusociality has evolved independently in several groups of Hymenoptera.
• Castes of ants: worker, soldier, alate (winged reproductive animals), queen.
• Only females are workers, all males are reproductive.
• All Hymenoptera are haplodiploid - males are haploid (eggs unfertilised), females are diploid.
• Workers' relatedness is 0.75 to sisters and 0.25 to brothers. Females are workers because they are more interested in rearing sisters (r=0.75) than their own offspring (r=0.5).
• Workers preferentially raise sisters, resulting in a theoretical 1:3 male/female ratio, although queen lays male/female eggs in a 1:1 ratio.
• Male/female ratio is indeed 1:3 in eusocial Hymenoptera, but in solitary Hymenoptera and in slave-making ants (where workers are unrelated to offspring) it is more like 1:1.
• At a 1:3 sex ratio, payoff for a worker is same as for a queen - so no bias towards eusociality?

Eusociality in Isoptera (termites)

• Termites are diploid, and both sexes become workers.
• King and queen are highly homozygous (inbred), because reproductives are produced not by founding king and queen but by their offspring (brother-sister mating).
• Relatedness of offspring is 0.5, but relatedness of siblings is nearly 1, favouring eusociality.

Eusociality in mammals

• Naked mole rat and Damara mole rat colonies: reproductive, guarding, digging and foraging castes.
• Colony is xenophobic and highly-inbred (mean r=0.81), favouring eusociality.
• System has arisen in harsh environments where the colony requires a large workforce.
• Some other mammals approach a eusocial state (e.g. badgers, where dominant female exerts pheromonal control over the others to prevent them breeding).

Foraging

Assumptions of foraging theories

• Decision assumptions - what do we think the animal is deciding?
• Currency assumptions - how does it evaluate the decision (e.g. energy maximisation)?
• Constraints - intrinsic (discrimination ability), extrinsic (characteristics of the environment)
• Conventional constraints: exclusive search and exploitation (can only do one thing at a time), sequential random encounters, complete information.

Functional responses

• Holling's disc equation: P_e = (a' N T) / (1 + a' N T_h) where P_e is number of prey eaten, a' is searching efficiency (attack rate), N is prey density, T is total time, T_h is handling time.
• Intake rate, P_e/T = (a' N) / (1 + a' N T_h)
• At high prey densities, intake rate is inversely related to handling time (P_e/T --> 1/T_h).
• Holling Type 2 functional response is hyperbolic curve (e.g. in parasitoid wasps).
• High a' produces steeper curve than low a'. Damselfly nymphs: larger prey have higher handling time, lower attack rate, shallower functional response curve.
• Type 1 functional response: linear increase in prey intake up to a maximum.
• Daphnia filter-feeding on yeast: Type 1 functional response due to simultaneous search and handling.
• Type 3 functional response: sigmoid curve (sometimes seen in parasitoid wasps due to effect of probing time, which increases with larger prey).

The prey model

• Energy intake rate, R = E / (s + h) where E is energy content of item, and s + h are search and handling times.
• Prey item should be taken if it would improve average energy intake: E_i/h_i > mean E/(s+h)
• If food is easy to eat but hard to find, predator should be indiscriminate. If food is abundant but hard to handle, only high-profitability foods should be eaten.
• Poor environments (high s) should promote less specialised diets than rich environments.
• Overall energy intake rate for a group of prey items: R = Sp_il_iE_i / (1 + Sp_il_ih_i) where is l is encounter rate and p is probability of prey being attacked - this is similar to the disc equation.
• Prey types should either always be taken or not (zero-one rule).
• Prey types should be ranked by profitability (e/h) and added to the diet in rank order, until profitability of next item is less than weighted average profitability of previous items.
• Inclusion of an item in diet depends on its profitability but not on its encounter rate.
• Crabs eating mussels: prey choice not quite as expected - a few low-profitability items taken.
• Possible reasons for this violation of theory: discrimination errors, long-term learning, prey parameter variation, runs of bad luck, simultaneous encounters, averaging across individuals.
• Great tits eating mealworms: encounter rates of different-sized prey manipulated experimentally - results closely match prey model predictions, but still a few inferior prey items were taken.
• Lions eating herbivores: observed diet closely matches predictions of prey model.

Marginal value theorem

• Gain curve (residence time in a patch, t_i, against cumulative gain, G) can be plotted on a graph - optimum residence time (maximum G/t) is point at which a tangent drawn from time 0 touches the gain curve.
• If patches differ, the optimum strategy is to treat each as if it were the 'average patch'.
• Travel time is positively related to time spent in patch (in great tits, paper wasps, starlings, etc).
• Patch depression: intake rate declines as exploitation of a patch proceeds.
• Possible patch use rules: leave after catching n prey, leave after t seconds, leave after g seconds of unsuccessful search (this 'giving-up time' rule produces results that closely approach the optimal solution), leave after instantaneous intake rate drops below a critical value.

Diet balancing

• Amount of Food A eaten can be plotted against amount of Food B eaten, and lines drawn representing limits (e.g. minimum energy requirement, maximum foraging time).
• The energy maximising solution (e _max) will be at one corner of the feasible area; the time-minimising solution (t _min) will be at another.
• Moose, whose normal diet is twigs, also eat pondweed because they need sodium. Like many herbivores, they appear to be energy-maximisers, but there are doubts about the validity of this conclusion.

Why animals live in groups

Finding food as a group

• Finding food: active recruitment (e.g. honeybees), cue in on successful foragers (local enhancement - e.g. Japanese macaques), follow successful foragers from roost (e.g. common terns).
• Catching difficult prey (e.g. hyenas hunt in larger groups for zebra than for wildebeest).
• Lions: when food is scarce, solitary females do better than groups of 2-4, but groups of 5-6 do best.
• Harvesting renewing food - more efficient to do it as a group than to randomly compete (e.g. hummingbirds at a feeder, where competitive losses are balanced against timing losses).
• Preventing interference by defending a territory.
• Costs of group feeding: prey disturbance (e.g. redshanks on seashore), problem of divisibility of food source (house sparrows hog indivisible bits of bread but share crumbs).

Avoiding predation as a group

• Vigilance - when predators rely on surprise, many pairs of eyes are better than one.
• Redshanks threatened by peregrines: individual scanning rates are lower when in a larger group.
• Why should animals do their fair share of scanning? Cheetahs chase less vigilant gazelles.
• Assumption of collective detection. If one sparrow flies off (when experimenter throws something at it) the others don't usually follow, but if two sparrows leave at once the whole flock takes flight.
• Response affected by distance to cover (the sparrows fly off more readily when far from cover).
• Edge effect - those on the edge of a group are more vulnerable and can see better - they should be more vigilant. True in starlings, but not in redshanks (which are threatened by aerial predators).
• The Selfish Herd: cost (higher probability of attack) < benefit (lower probability of capture)
• Confusion effect - difficult to pick out an individual in a herd (explanation for zebra stripes).
• Alarm signalling - why draw attention to yourself? To protect relatives, or get lost in chaos.
• Dilution effect - large groups overwhelm predators (e.g. Adélie penguins entering water).
• Encounter effect - decreased probability of detection by predator when animals are together.
• In redshank groups larger than 30, probability that at least one individual has its head up decreases. They may be relying on dilution effect.
• Living in groups increases conspicuousness, but this cost tails off with increasing size, so some animals may live in massive groups (e.g. monarch butterflies).
• Ostriches and cichlids steal each others' young to create larger groups.
• Group sizes tend to be larger than optimal, because an individual's alternative (going it alone) is worse.

Predator-prey interactions

Avoiding detection by predators

• Prey may avoid detection by: immobility, crypsis, confusion, exploiting perceptual limitations.
• Apostatic (frequency-dependent) selection - most predators prey preferentially on common prey types (e.g. chicks fed green and brown pellets), due to formation of search images etc, so rarity is advantageous. Some prey may appear rare through polymorphism (e.g. snails).
• Probability of detection increases with viewing time - longer for more cryptic prey.
• Guppies have smaller and fewer spots when under high predation pressure.
• Crypsis depends on context. New Zealand birds are drab because their predators can see in colour.
• Confusion: unpredictable movement, extreme abundance, flash colouration, polymorphism.
• Sensory and perceptual limits: minimum distance for pattern element or colour, flicker fusion, private wavelengths, sealed shells to prevent chemical leakage.

Avoiding attack by predators

• Avoid identification: masquerade as an inedible object, confusion, aposematism (warning colouration), mimicry, honest signalling of unprofitability (e.g. stotting in gazelles).
• In Müllerian mimicry both species are unpalatable; in Batesian mimicry the mimic is harmless.
• Mimicry is not just visual (e.g. ultrasound in distasteful moths at risk from bats).
• Predator learns warning colouration more effectively at high prey densities - strong stabilising selection, leading to monomorphism.
• How does distastefulness evolve? Prey surviving attack? Kin selection? Synergistic selection?
• Predators learn a conspicuous colour pattern quickest (chick-eating-pellet experiments again).
• Aposematic patterns should be simple, repeating and all-over.

Arms races and co-evolution

• In parasite-host systems there is very specific adaptation (e.g. cuckoos).
• True co-evolution is rarer in predator-prey relationships, but there are some examples - night-flying mantids, unlike most mantids, can hear bat echo-location frequencies.
• In general, groups of predators co-evolve with groups of prey: as more shell-breaking predator families evolved, unshelled gastropods declined and gastropods with good apertures increased.
• Evolution tends to be faster in prey than in predators due to life/dinner principle (the rabbit has more to lose than the fox!), larger numbers of prey and shorter generation times.

Avoiding approach of a predator

• Mode of fleeing: outrun predator, sprint to cover, use different style of motion (e.g. fly away).
• Behave unpredictably, or use startle, bluff or threatening behaviour.
• Encourage premature attack (e.g. by stotting).

Avoiding capture by a predator

• Avoiding subjugation: strength to escape, physical toughness (e.g. ticks), mucous or slime, autonomy of body parts (e.g. lizard tails, starfish arms), head hiding or false heads, spines or other structures, adhesives (e.g. black widow silk), jaws and claws, noxious tastes, poisons, stings, group defence, resistance to venom (e.g. California ground squirrels and rattlesnakes).
• Avoiding consumption: safe passage through gut (e.g. molluscs), emesis (triggering vomiting - e.g. Monarch butterflies), being poisonous (may only benefit prey through kin selection!).
• It is in prey's interest to interrupt predation as early as possible: at later stages, predator is increasingly close, more energy is required to stop it, and predator is less willing to give up.

Life histories and reproductive decision-making

Trade-off between current and future reproduction

• Can be studied using brood-manipulation experiments.
• After artificial brood increases, 5 out of 14 studies found decreased local survival of parents (mortality? emigration?); 8/14 found decreased later fecundity (physiological impairment?).
• Effect of brood increase on parents is not compelling, but increase in chick mortality resulting from increase in brood size is dramatic.

Timing breeding to coincide with peak in food supply

• Variations in timing of breeding are retained in population as a balanced polymorphism by year-to-year variations in timing of food supply.
• Earliest breeders are typically most productive - average bird appears to behave sub-optimally. This is explained by individual optimisation. Trade-off between fewer superior young earlier and more inferior young later - depends on quality of territory/individual.

Variations in parental risk-taking and investment

• Adults of longer-lived, less fecund species should be less willing to put themselves at risk for their offspring. Nuthatches belonging to species with higher fecundity/lower survival respond more strongly to egg predators; other nuthatch species respond more strongly to adult predators.
• Males should invest more in their young if monogamous.
• In polygynous species, females in good condition should favour sons. Feral horses in New Zealand: females in good condition allow sons to suckle longer; mares in good condition at conception give birth to more sons; those in poor condition produce more daughters.
• Optimum investment in offspring during pregnancy is higher for foetus than for mother.
• Pregnancy sickness in humans: benefits foetus (by reducing exposure to toxins) but harms mother.
• Manipulation of mother by foetus - foetus secretes placental lactogen, which soaks up insulin so blood glucose rises, mother responds by overproducing insulin - may lead to diabetes.

Role of behavioural ecology in conservation

• Small populations - are they at particular risk (through reduced vigilance/group defence)?
• Sex ratio - might additional feeding (e.g. kakapo), or hunting, alter the ratio?
• Understanding mating system. Sperm whales: originally thought to have harem defence, so most males were superfluous and it was OK to kill them. Now shown not to be the case - a few older males move between groups of females while young males remain at high latitudes.
• Predator-prey interactions - are ecotourists perceived as predators?
• Exploitation (e.g. Atlantic cod: at low densities, fish congregated in preferred areas, masking decline).
• Reducing predation - killing off predators (e.g. introduced mammals in New Zealand) is often ineffective and problematic. Habitat can be manipulated to reduce predation. Deterrents can be used (e.g. electrified dummies to reduce tiger predation on humans).
• Conservation practices may increase predation (e.g. removing rhino horns to deter poachers).

Animal communication

• In 'true communication', both sender and receiver benefit (on average) from the exchange.
• The receiver's response to a signal affects the sender's fitness as well as its own.
• Information exchange may occur without signalling (e.g. 'local enhancement' at feeding sites).
• Communication may be interspecific (killdeer plover feigns broken wing to lead fox away from nest).
• 'Amplifiers' may assist in signalling (e.g. dark markings to highlight size of fish).

Signal evolution

• Levels of analysis: causation, survival value (function), evolution, ontogeny (development).
• Ways of analysing displays: motivational, interactional, economic.
• Different channels of communication (chemical, auditory, visual, tactile) have different properties (range, rate of change, passage round obstacles, locatability, energy cost).
• Model of signal evolution: sender associates incipient signal and condition, receiver perceives signal against background, receiver associates incipient signal with condition, receiver associates updated information with decision, receiver responds. Feedback loop leads to ritualisation and refinement of signal.
• Signals can jump from one system to another. Macaques: females present sexual swellings on rear to males to show receptivity. Males use this presentation posture to each other as sign of appeasement. In some species, male rear mimics female rear to enhance this signal.
• Ambiguity - tits displaying at territorial boundaries: a head-up display has an equal probability of leading to approach/attack or to fleeing; head-down birds are more likely to approach/attack.
• Receiver's actions can be important - pigeon guillemots give hunch-whistle displays when defending their territories: if intruder responds by sitting, owner may sit too, but if intruder walks, owner is likely to hunch-whistle again.
• Sender pre-adaptations: intention movements (e.g. sky pointing in gannets before leaving nest), ambivalent behaviour (e.g. forward threat posture in black-headed gull), protective responses (e.g. primate facial expressions), autonomic responses (e.g. urination leads to scent marking, heavy breathing to vocalisations), displacement activities (bits of out-of-context behaviour - e.g. preening in duck courtship), redirected attack (e.g. grass-pulling in aggravated herring gulls).
• Receiver pre-adaptations: lizard vision is attuned to jerky movement, leading to head-bobbing displays; male water mites smell like food to attract females.
• Environmental selection pressures: eavesdropping by predators (e.g. snakes following mouse scent trails), physical structure of environment, influence of other signallers.
• Great tits alarm calls: mobbing call (intended to be heard by predator) spans wide range of frequencies; "seet" alarm call (not intended to be locatable by predator) is a pure tone, at a frequency that great tits can hear better than sparrowhawks (predators) can. This creates convergent evolution among passerine birds (different prey species have similar alarm calls) and arms races between predators and prey.

Social organisation

• Social organisation is a dynamic phenomenon, and may vary between populations of the same species (langur monkeys live in larger groups where habitat is richer).
• Social organisations reflect adaptive responses to selection pressures acting on individuals.
• Females are the limiting sex; female distribution depends on ecological factors. These influence distribution of females, size and characteristics of female groups, and interactions with males (need for paternal care, feasibility of mate guarding, etc). These in turn influence distribution and behaviour of males. Social organisation arises from these interactions.

Describing social organisation

• Population structure (demography). Rhesus monkeys: in larger groups, average relatedness is lower and infants keep closer to their mothers - group size influences maternal behaviour. Large monkey groups are likely to fission (split) along matrilineal kinship lines.
• Spacing behaviour: inter-society dispersion, or dispersion of solitary animals.
• Social system: behaviour within societies.
• Mating system: monogamy, polygamy, polygyny, etc.
• Changes with time. Red deer become aggressive and territorial during rutting season. Tits and chaffinches are territorial during breeding season but form flocks later in the year.

Spacing behaviour

• Home range: the area used by an animal during routine daily activities.
• Home ranges may overlap extensively when costs of defence are too great (e.g. in baboons).
• Territory: fixed area from which intruders are excluded by advertisement, threat or attack.

Animal groups

• Solitary (e.g. bears, cats other than lions).
• Aggregations: animals are drawn to environmental resources together.
• Anonymous groups (e.g. migrating herds of ungulates, flocks of birds).
• Societies: usually have restriction of membership (same species), limited membership (same individuals), definite attraction between individuals, contrast between inter- and intra-society communication, division of labour, co-operation, synchronisation of activities.

Baboon societies

• Four different species (or subspecies) of 'common baboon' (Papio) are widely-distributed in sub-Saharan Africa; their societies differ from those of the 'sacred' Hamadryas baboon.
• Common baboons live in multi-male/multi-female groups (no 'bachelor groups').
• Hamadryas baboons have a harem group structure: a troop consists of an association of many 'clans', comprising harems and bachelor males.
• Common baboons will learn Hamadryas baboon ways if introduced to a Hamadryas troop.
• Common baboons living on the savannah have large, overlapping ranges; those living on forest edges (where food is more abundant) have smaller ranges and are more territorial.

Social hierarchy in baboons and macaques

• Dominance: a consistent outcome of social interactions in favour of the same dyad member, and a default yielding response of its opponent rather than escalation.
• In primates, males usually outrank females. Male and female ranks are drawn up separately.
• In common baboons: male ranks are less stable than female ranks (due to movement of males between groups), male ranks are based on age, female ranks are based on matrilineal kinship.
• High rank doesn't necessarily equate to a high level of aggressiveness.
• Supplanting: one baboon takes over another's food source. Corms (swollen bulb-like stem bases) are harder to find than grass, but supplanted more because they have higher nutritional value. Supplanted baboon is less likely to resume feeding on same food type if it is corms. Proportion of supplants involving grass (vs. corms) is higher when baboons are close in rank - they take the opportunity to reinforce rank positions. Baboons know each others' ranks.
• Basic rank: judged on the basis of dyadic interactions (between two animals).
• Dependent rank: depends on the behaviour of others (e.g. relatives who intervene in fights).
• Dependent rank may become assimilated into basic rank.
• Females support relatives (altruistically) in fights; the risk they are willing to incur depends on relatedness. Animals will intervene readily in fights with low-ranking individuals (low risk), but only intervene in fights with high-ranking individuals if a close relative is involved.
• Protective threat position: juveniles may put themselves beside an adult and threaten another juvenile, so any reciprocal threat also annoys the adult.
• Sub-adult male may grab a baby and use it as an excuse to approach an adult male - they can enhance their rank by associating with dominant males.
• Rhesus macaques scream to recruit assistance in fights - type of scream depends on who is attacking and whether or not there has been contact.
• Facial expressions are also used - different expressions can be given to different individuals.
• Higher ranking females have greater reproductive success (breeding earlier in life, shorter inter-birth intervals, higher fecundity, lower infant mortality) due to better nutrition.
• Male rank is also correlated with reproductive success (although this is hard to measure due to changes in rank and uncertainty of paternity).
• Daughters inherit maternal rank.
• Principle of young ascendancy: younger siblings come to outrank older ones due to preferential treatment by mother (younger siblings have a higher future reproductive value).
• Among animals in captivity, dominance hierarchies may be badly distorted.

Practicals and self-teach work

The amorous Gammarus

• In freshwater shrimp Gammarus pulex, male guards female for several days prior to copulation.
• Number of eggs produced by female is correlated with body size.
• 'Giving up time' is time taken to separate when pre-copula pairs are taken out of water.
• Body lengths are normally distributed; cube roots of giving-up times are normally distributed.
• Male and female sizes within mating pairs are strongly correlated.
• Male and female sizes are correlated with giving-up times, but not statistically significantly.
• Both males and females prefer larger mates.

Sexual dimorphism in primates

• Sexual dimorphism (male/female mass) is significantly correlated with male competition.
• Arboreal/terrestrial species are more sexually dimorphic than purely arboreal species.
• Frugivores are more sexually dimorphic than folivores or frugivore/insectivores.

Feeding behaviour in locusts

• Bite mass is larger on cabbage than on barley or grass.
• Food intake rate (bite mass × frequency) is higher on cabbage and grass than barley.
• Cabbage was the preferred food (measured by time spent and food ingested).

Foraging behaviour - the disc equation

• 'Forager' has Type 2 functional response (hyperbolic curve) due to finite handling time.

Statistics

Methods of recording behaviour

• Sampling rules: ad libitum, focal (monitor particular individuals), scan (do a rapid census).
• Recording rules: continuous (record frequency or duration of events), time sampling (record behaviour at specific times), one-zero (record whether an event happened during time period).

Levels of measurement

• Nominal: classification into labelled categories with no natural order (e.g. male/female).
• Ordinal: ordered but differences between values are not important (e.g. rankings).
• Interval: ordered, constant scale but no natural zero (e.g. temperature in °C).
• Ratio: ordered, constant scale with natural zero (e.g. age, size).

Analysing differences in dependent variables measured on a numerical scale

• With independent samples measured under two conditions, use 'unrelated' 2-sample t-test (parametric) or Mann-Whitney U-test (non-parametric).
• With paired samples measured under two conditions, use 'matched' paired-sample t-test (parametric) or Wilcoxon matched pairs test (non-parametric).
• With independent samples measured under three or more conditions, use one-way Analysis of Variance - ANOVA (parametric) or Kruskal-Wallis test (non-parametric).
• With paired samples measured under three or more conditions, use repeated measures one-way ANOVA (parametric) or Friedman test (non-parametric).
• Use two-way ANOVA if there is more than one independent variable.

Analysing trends

• If x and y values are both measured, and neither is imposed by experimenter, use Pearson correlation (parametric) or Spearman rank order correlation (non-parametric).
• Use regression analysis (parametric) if x values are imposed and y values are measured.

Analysing differences in dependent variables measured on a categorical scale

• Use chi-square goodness of fit test or G-test to measure goodness-of-fit between observed and predicted values.
• Use chi-square contingency test or G-test contingency test to test for association between two factors.

More notes and essays

© Andrew Gray, 2004