Phase change in locusts

Question 1

Characteristics of gregarious desert locusts - brain

Answer 1

Body mass 20% smaller
Brain mass 28% bigger
No difference in antennal lobes - 50% bigger calyx and mushroom bodies (allometry)
30% bigger central complex
No difference in lamina-20% bigger medulla-30% bigger lobula (allometry)
30% bigger central complex (mosaic)

Question 2

What are gregarious vs solitarious locusts?

Answer 2

Locusts are just grasshoppers that can phase change
10 known locusts, widely scattered across grasshopper family (i.e. phase change evolved several times independently, convergent evolution)
Locusts reared on their own become solitarious, reared in a crowded group become gregarious
Solitarious desert locust found across Saharan Africa (its recession zone), expand further north and south when swarming & gregarious (to affect 20% of earth’s surface)

Question 3

Phase change basics - what is it, why does it happen

Answer 3

Form of phenotypic plasticity, occurs in response to population density (i.e. no external environmental factor)
Locusts occur in regions with unpredictable/infrequent rainfall - after rain, there are more locusts (because more vegetation), then when the vegetation dies back the increased numbers condense into a smaller area, and phase change then swarm.
Can happen at any point in lifetime, or not happen at all. Behaviour changes over a few hours to days, colour begins overnight but can take several generations to fully manifest, full suite of morphology takes several generations
Gregarious female can transmit phase to offspring by secreting substance onto eggs (washing the foam off them leads to solitarious offspring), so it must be epigenetic. This is how the full transition can take multiple generations

Question 4

Comparing solitarious and gregarious desert locusts (10)

Answer 4

Solitarious have cryptic colouration - green for camouflage as a nymph, brown as an adult.
Gregarious have aposematic colouration - black and yellow as a nymph, yellow as adult

Solitarious walk slowly, creeping
Gregarious walk faster, with high-stepping gait

Solitarious are crepuscular
Gregarious are diurnal

Solitarious have a restricted diet
Gregarious eat a wider range of plants, including some poisonous (that protect against predation)

Solitarious have low fat deposits
Gregarious have high fat deposits

Solitarious have larger eyes and antennae (i.e. distant receptors)
Gregarious have more mechano- and taste receptors all over their body (i.e. detect things close to them)

Solitarious live up to 2 years
Gregarious live around 6 months

Solitarious have longer wings and hind legs
Gregarious are more compact

Solitarious tend to avoid other locusts
Gregarious live in groups

Gregarious groom more frequently and have stronger immune systems

Question 5

Measuring phase

Answer 5

Using binary logistic regression, P compresses frequency of grooming behaviour, time spent still, time spent adjacent to locust-containing arena, walking speed into one measure of gregarity

Question 6

Behavioural gregarisation (desert locust)

Answer 6

Behavioural transition (which occurs first, is the most labile aspect of the change, and is reversible) encourages further phase change, by keeping the locust close to others. 
Gregarious phase change is triggered either by mechanosensation on hind legs [shown by tickling hind femura for four hours with a paintbrush, even if it's never seen another locust in its life it behaves fully gregarious], or visual and olfactory stimulation together ['cephalic pathway']
Patterned stimulation of a metathoracic nerve also induces gregarisation (simulates mechanosensory stim)

Question 7

Finding a chemical for gregarisation - experimental, desert locust

Answer 7

High performance liquid chromatography told us which substances were more prevalent in long term gregarious vs 3rd gen solitarious brains and thoracic ganglia. 11 out of the 13 chemicals analysed were different. But this can’t inform on function.
Then took LT gregarious and isolated for 24hrs, then 1st instar, then 1, 2 and 3 generations. Then crowded for 4hrs, then 24hrs, and so on.
Biggest changes occurred in first day of isolation and first day of crowding. The only change that occurred in the first 4 hours of crowding (i.e. mediating initial behavioural change) was serotonin, dramatically increased. Note that after 24hrs, this peak had dropped, so it’s involved in induction not maintenance.
Positive correlation was found between P and 5-HT in thoracic ganglia (not necessarily on duration of crowding)
5-HT antagonist injected into thoracic ganglia prevented gregarisation. So 5-HT is necessary
Applying 5-HT or agonists to thoracic ganglia increases gregarity of behaviour (though not full gregarious). So 5-HT is sufficient.

Question 8

Behavioural solitarisation (desert locust)

Answer 8

A newly gregariously behaving locust (that’s only been crowded for up to a day) becomes solitarious after 4 hours of isolation
A locust that’s been gregarious all its life takes more than 96 hours to become solitarious

Question 9

Finding pathway for gregarisation (i.e. specific protein pathway) desert locust

Answer 9

Serotonin shown to be sufficient and necessary
PKA inhibitor prevented gregarisation, but did not affect already gregarious locusts
RNAi of PKA catalytic subunit reduced gregarisation
Subjects with the least PKA catalytic subunit were the least gregarious
RNAi of PKA regulatory subunit increased gregarisation

So 5-HT –> Gs –> cAMP –> PKA
PKA catalytic unit perhaps interacts with CREB, transcription factor

Question 10

Finding neurons for gregarisation, desert locust

Answer 10

Immunofluorescence for 5-HT was used in thoracic ganglia
A subset of serotonergic neurons increased their production during gregarisation, especially in meso and metathoracic ganglia.
Somata staining intensity differed between phases and treatments. Some were more stained in gregarious (LT or ST) but didn’t respond to stimuli, some responded to sight and smell, some responded to touch. Some responded to multiple gregarising stimuli, and were more brightly stained in locusts subjected to the short term gregarising stimuli than LT gregarious locusts. The latter two groups, and especially the last, are the ones that could mediate the change.

Question 11

Finding genetic correlates of gregarisation, desert locusts

Answer 11

Locusts have genomes almost three times larger than ours
Lots of transposable elements and pseudogenes, makes genomic studies difficult
The migratory locust has a published genome, but still needs annotation and assembly

In an Expressed Sequence Tag study, 100 genes were more expressed in LT gregarious, 114 in LT solitarious (about 1% of genes present on microarray)
This is probably an underestimate, because the entire nervous system was analysed (so small changes could be hidden in noise), and they looked at mRNA so high turnover proteins could have been masked

40% of those hits have been annotated
Most are in metabolic pathway - solitarious locusts had genes to protect against oxidative stress, and for anabolic turnover (probably because they live longer). Also juvenile hormone-binding protein and hexamerins
–PENICK et al 2012 - juvenile hormone is used to mediate differentiation into different castes in ants, those with more nectar have more JH and become queens. Soldiers have higher threshold for JH –
–Martins et al 2010 - hexamerin genes in honey bee are upregulated by application of JH. Two of them showed caste-specific expression, highest in workers–
Gregarious had visual and olfactory genes upregulated, and heat shock proteins (because stressful environment), and genes for neural growth and renewal.

Question 12

Descending contralateral motion detector (as an example of neural effect of phase change)

Answer 12

Large diameter visual interneuron, carries info from eye to motor centres of legs and wings in thoracic ganglia.
Monosynaptic connection to fast extensor tibiae (FETi) in metathoracic ganglion, which controls jumping

Responds to small objects moving across field, but strongest response to looming objects, coming directly towards - i.e. near exponential increase in angle subtended across the eye
Spike frequency increases as the object approaches
Habituated with repeated stimuli, strong attenuation by the 9th stimulus.

Question 13

Differences in DCMD behaviour between gregarious and solitarious locusts

Answer 13

Gregarious have a stronger response across most of the field, but similar in peripheral regions
Solitarious habituate much more than gregarious
Solitarious habituate equally strongly across the field, gregarious have specifically reduced habituation in a central zone above the eye equator (where most other locusts would probably be)
Each solitarious DCMD spike produces an EPSP twice the amplitude of gregarious

SO, in gregarious, DCMD responds more strongly to a stimulus (especially after repeated presentation), but FETo responds less strongly to DCMD spikes.

Question 14

Differences in FETi response between gregarious and solitarious locusts, synaptic strength effects

Answer 14

Due to summation, in unhabituated locusts, the compound EPSP in FETi is the same in gregarious vs solitarious.
I.e. lower frequency in solitarious is balanced by greater synaptic strength onto FETi.
However, because solitarious locusts habituate, after a few presentations the compound PSP decomposes, whereas the gregarious response is just as strong.

There’s even more happening though - the gregarious DCMD-FETi synapse shows much stronger facilitation than in solitarious locusts, ensuring that the compound PSP is robust. This also makes the peak of the compound PSP occur earlier in gregarious.
There’s also improved signal-to-noise ratio, due to the lower initial synaptic strength in gregarious.
Resistance to habituation, increased signal-to-noise ratio and earlier rise to maximum in motor neurons helps gregarious locusts whose entire visual field is dominated by other locusts.

Question 15

DCMD and circadian pattern

Answer 15

Solitarious locusts show a normal sine-wave with peak at expected dusk
Gregarious have more complex pattern, peaking mid-afternoon, staying high for rest of the day

These are controlled internally, independent of environmental illumination.

So the neuron’s activity has adapted to the different phase and its activity pattern

Question 16

Migratory locust

Answer 16

Major pest in Asia, but widely distributed across world - only place it’s not found is the recession zone of the desert locust!
Different subfamily to desert locust
Its own recession region is wetland, factors that drive swarming not as clear as in the desert locust.

Question 17

Australian plague locust

Answer 17

Major pest in Australia (government commission to control it)
Half the size of the other species
Throughout the Australian outback (not found in most fertile areas or cities)
Unpredictable drought/rainfall fluctuations, so similar to desert locust
Not as dramatic-looking, solitarious and gregarious are similar in colour, gregarious slightly paler.
They also have humidity and temperature-related colour differences
Same subfamily as migratory locust, but not v closely related

Question 18

Behavioural phase change in australian plague locust

Answer 18

Attraction to other locusts and degree of activity change simultaneously, rather than the clear positive feedback loop seen in desert locusts (where attraction causes the other behavioural changes)
So perhaps one overarcing mechanism
Initial rapid phase of gregarisation takes 4hrs, then slow phase takes about three days (this doesn’t occur in desert locusts)
Same pattern seen in solitarisation.

Question 19

Comparing rates of phase change between species

Answer 19

Gregarisation:
Desert locust - 4hrs [serotonin in thoracic ganglia]
Australian plague locust - fast phase 4hrs, slow phase 3days [octopamine]
Migratory locust - More than 3 days (we don’t see any behavioural change at that point) [dopamine]

Solitarisation:
Desert locust - At least a few days [???]
Australian plague locust - fast phase 4 hrs, slow phase 3 days [5-HT increases for 1hr then decreases for 3 days]
Migratory locust - 4hrs [serotonin increases over 32hrs despite behavioural change being complete. Agonists had no effect]

Question 20

Comparing stimuli for phase change

Answer 20

Gregarisation:
Desert locust - mechanosensation on hind legs, or visual and olfactory stimulation
Australian plague locust - mechanosensation on antennae

Question 21

Visual stimuli importance in gregarisation in APL

Answer 21

APL - if while isolated they can see one other locust, they’ll stay gregarious. Some evidence suggests the more locusts they can see, the longer they’ll stay gregarious

A recently solitarised locust just has to see a locust to be re-gregarised
Even a dead locust was sufficient stimulus, though a bead was not - so it’s shape based, not movement
Only works over distances shorter than 15cm, due to angle subtended on retina (shown by using different sizes of animal at the same distance)

Question 22

Olfactory stimuli importance in gregarisation in ML

Answer 22

Gregarising stimuli still largely unknown, but a microarray and qPCR showed chemosensory proteins more highly expressed in gregarious locusts
RNAi of CSP3 in gregarious nymphs caused them to avoid other locusts and avoid locust volatiles in a Y maze

Takeout gene is an olfactory binding protein, more expressed in solitarious locusts
RNAi of takeout made solitarious locusts spend more time near group, and abolished preference in Y maze
Possible that this is just reducing their sense of smell generally.

Question 23

Finding a chemical for gregarisation (APL)

Answer 23

All the major biogenic amines found in greater concentrations in gregarious than solitarious, but increase over different timecourses
5-HT peaks at 4hrs, so can’t be responsible for slow phase
Octopamine tracks behavioural change most closely - but this is just a correlation

Question 24

FInding a chemical for solitarisation (APL)

Answer 24

None of the biogenic amines closely follows pattern of behavioural change
Serotonin increases within 1 hour then decreases over 5 days
Octopamine remains similar to gregarious levels, even when animal has become fully solitarious in character

Question 25

Finding chemicals for phase change in ML

Answer 25

None correlate with behavioural change
Crowding doesn’t change serotonin
Conflicting results for serotonin

serotonin increases over 72 hours of isolation, which is way beyond manifestation of behavioural change. It may enhance rate, but clearly not causal

Dopamine has the closest link
Gregarisation:
brain titres increase during crowding, as does D1R, and upregulation of DA synthesis pathway genes (henna, vat1 and pale)
Inject gregarious locusts with DA antagonists, or interrupt signalling using RNAi, you get solitarious behaviour
Injecting DA agonists gives acute change towards gregarious, and enhances crowding-induced gregarity

Solitarisation:
Isolating gregarious gives sudden drop in DA titres (after 1hr), and rise in D2R expression
Isolating gregarious with D2R antagonist protects gregarious phenotype

Question 26

Phylogeny of phase change

Answer 26

Only one Old World Schistocerca (‘bird grasshopper’), though most members of its subfamily are old world.
50 American species, Old World Hypothesis says desert locust got across the Atlantic by hopping from raft to raft of dead locusts from the front of the swarm, eating them for energy (in 1988, a swarm of desert locusts got from west africa to west indies in this way)
But then why hasn’t desert locust diversified in Old World? Why would its New World offspring speciate extensively, but itself not evolve much?
And what about the New World genus Halmenus?
New world schistocerca are very diverse, about half have become obligatory solitarious
The rest show some aspects of phase change - maybe just colour, maybe just behavioural changes, maybe both
Can use these to look at genetic basis of phase change

Song suggests that Schistocerca diversified in the New World, then one species made it across to Africa. This was also the original hypothesis.

Question 27

Searching the egg foam

Answer 27

Miller et al 2008 used HPLC on egg foam from gregarious desert locust mothers and applied increasingly specific fractions to solitarious eggs, then behaviourally assessed the hatchlings. They narrowed it down, then used NMR spectroscopy to identify the component conferring the phase.
They found that an alkylated L-DOPA analogue triggered gregarisation in a dose-dependent way. However, it was expressed in equal amounts in solitarious and gregarious foam - instead, it seems another compound may inhibit it in solitarious foam. They found a compound that was molecularly similar (so could compete), and elevated in solitarious foam. If synthetic inhibitors similar to this compound could be developed, they could prevent swarming without harming the insects.

Question 28

Phylogeny and evolution of phase change

Answer 28

Song, 2005 - It was originally assumed that phase change arose through convergent evolution. This is flawed because

1) Phase is not binary, it’s a composite character. Some species show some aspects of phase but not others, suggesting that they’re a collection of different reaction norms that don’t necessarily co-evolve.
2) distribution of locusts within each subfamily of Acrididae isn’t random - e.g. 4 within Schistocerca. This suggests it evolved from a common ancestor.

Song proposes that swarming developed separately from and later than phase change - whilst aposematic colouring has been shown to be adaptive in dense desert populations, and density-dependent immune enhancement has been shown, there are plenty of grasshoppers that phase-change without swarming, and plenty of different animals that swarm without phase change. In no case were locust species basal to sedentary species in this phylogenetic analysis

Remember phenotypic plasticity can retard natural selection since it makes the same genotype adaptive in multiple environments. So some grasshoppers may have hidden phase change capabilities, and just haven’t had sufficient trigger to swarm.