Case Study 1 - Phenology Match-Mismatch

MCQ revision

Phenology

Timing of recurrent life history events, for example: the emergence of butterflies

Why is Phenology a popular research are in the context of climate change?

Events are getting earlier

Different populations are changing timings at different rates, so the interactions between organisms are being altered causing match-mismatch phenomenon

Define the Match-Mismatch hypothesis

The match-mismatch hypothesis hypothesises that due to climate change, certain events such as butterfly emergence, are getting earlier, which means that populations that interact with butterflies (plants) are unable to do so at times that they normally would, altering their interactions => A consumer relies on a resource, if they mismatch their timings with that resource, there are negative implications for their survival/fitness

Provide an example to illustrate the term ‘match-mismatch’

an example is: If zooplankton populations peak earlier and fish do not alter their hatching times in response to this, there will not be enough zooplankton to eat when fish do hatch. As a result, there will not be enough food for fish and they die.

What elements need to be present for match-mismatch to happen

1. The consumer must be dependent on a time limited resource,

2. There must be a degree of asynchrony that varies from year to year, do that there is variation in mismatch,

3. the degree of asynchrony must be increasing,

4. this must have a negative impact on the consumer.

Thackeray et. al. showed that the conditions for mismatch are common, what did they find?

1. They found that phenology is getting earlier by 3-4 days per decade

2. They found that primary producers and consumers are shifting their phenology by a lot

3. they found that secondary consumers are shifting their phenology less, causing more asynchrony.

Thackeray et. al. also looked at how much phenology was shifting with the warmth of spring, what did they find?

They found that phenology shifts were primarily food chain based.

The great tit - caterpillar system is a good one to study match-mismatch, why?

• great tits rely on the winter caterpillars for feeding of their chicks

• they need to predict when the caterpillars will be at highest abundance 30 days before hatching

• this is so that they have enough caterpillars to feed their chicks 10 days after hatching, at their peak demand

• birds that are more matched tend to have higher fitness

• caterpillar hatching varies

• so the consumer is reliant on a time-limited resource

Why might habitat generalism act as a buffer?

If there is a metapopulation of great tits, made of communitues residing in birch forests or oak forests then even if one community is impacted by match-mismatch, another may not have this or it may be acting at reduced level. This would buffer the effects of match-mismatch at a metapopulation level, that is, the number of chicks that survive overall may remain constant at a metapopulation level rather than decreasing as it might in a severely mismatched sub population.

Design a study to find whether:
(i) habitat generalism acts as a buffer on mismatch. Test
whether the amount of asynchrony between great tits and
caterpillars varies and the impact of this on great tit fitness varies among
different woodland habitats

• Need to find a measure for fitness

• Need to find when caterpillars emerge and when chicks are laid in each forest and then do a correlation to measure match-mismatch

• catch birds, tag them, measure number of chicks fledgeing and parent survival to next season

• replications: 44 sites, 6 - 8 nest boxes per site,

• measure number of caterpillars by shaking a tree and counting droppings or by waiting for silk moths to descend and collect them when ½ the total have descended

Using 44 sites and 5-6 bird boxes for this study has what benefit and what drawback?

• Benefit: A lot of statistical power to estimate the
  relationship between drivers (e.g.,
  temperature/habitat) and among site
  variation in phenology or fitness.

• Drawback: imprecise estimate of mean values at each site

  
  

design a study to test: Does trophic generalism act as a buffer on mismatch? Test
whether the importance of winter moth caterpillars (or caterpillars
in general) to consumer diet varies among habitats

• Faecal metabarcoding: collect faecal samples, extract DNA, carry out PCR, sequence DNA, identify taxonomic units

• from this quantify the proportion of prey that belongs to any one taxon

Visser et. al. studied this system in the Netherlands:

• what did they measure?

• what did they find?

Visser et. al. measured

• selection differential - fitness vs lay date

• measured fitness by number of chicks fledging

Husby et. al. found that

• selection for earlier laying has become stronger over time and this coincides with warmer springs => this suggests that asynchrony between birds and caterpillars is increasing

• birds that do not adjust their timing are experiencing lower fitness

Charmantier et. al. studied this system in Oxford, what did they find?

they measured

• selection differential vs the interval btwn bird laying date and caterpillar peak timing

they found that

• warmer springs conferred stronger selection for earlier laying

• colder springs had a weaker selection pressure becauses caterpillars emerge later

• this is strong evidence that asynchrony impacts bird fitness

How did the netherlands team (Visser et. al.) measure caterpillar abundance? Pros and cons?

• they collected caterpillar droppings under trees and weighed the frass to estimate caterpillar abundance

  PRO: this is a direct measure of caterpillar feeding activity

  CON: labour - intensice + small scale

How did the Charmantier Oxford team measure caterpillar abundance

using the Half-Fall date

• winter moths descend on silks after feeding

• researchers collect them in pools of water

• the ½ fall date: when 50% of winter moths have descended

PROS: captures entire caterpillar population timing

CONS: Indirect estimate of caterpillar abundance

why is replication so important?

replications allow site effects to be taken into account.

replications provide statistical power

What did the faecal metabarcoding tell us?

They compared caterpillar availability to the proportion of caterpillar DNA reads in bird shit

Found that

• the winter moth was the most common prey

• no major differences in winter moth abundance from south to north

• nestlings consume mostly lepidoptera

• evidence that birds prefer winter moths but a switch to other caterpillars when necessary, ex: Pieridae

• brood success is linked to total Geometridae abundance, not just winter moths which suggests buffering against mismatch

• birds are not exclusive specialists

Propose an approach to seperate whether the relationship between timing and fitness is due to

• Timing aka is it better to breed earlier

• Individual quality aka do the best bird breed early and does their quality determine their fitness

do cross fostering aka egg-swapping btwn early and late-laying nests

• measure chick survival based on their new hatching time

• if individual quality: cross fostered chicks given to late parents should still thrive and those given to early parents should struggle

• if timing is key: early chicks given to late parents should do worse and late chicks given to early parents should do better

was the timing and fitness relationship due to timing or individual quality?

• Timing had a stronger effect than parental quality: chicks that hatched at the right time did better

• incubation cost differences may have confounded results (early birds forced to incubate longer)

Under the timing hypothesis, why would early laid chicks given to later parents have lower survival?

If its just better to be laid early, it would be due to optimal hatching conditions earlier on (resources, temperature, etc). If early birds were given to later parents then they will hatch later and parents will not have food for them and conditions will be sub-optimal

Why is replication key according to mr ally phillamore

• replication at the right level aka across sites and not just within them is critical for generalizable findings

• you cannot say birch tree habitats have higher mismatch if you only see it in one site

What can we do about the fact that observational studies are usually subject to confounding variables?

• we should try to control for these variabes in our analyses by design or by inclusion as covariates

• be careful about what we infer from correlational studies

• experiments are the best solution