Botany 2014 Boise, ID

Maternal effects reveal species niche dimensions Rachel M. Germain and Benjamin GilbertDept. Ecology & Evolutionary Biology, U. Toronto

“Why are there so many kinds of [species]?”-G. E. Hutchinson 1958

Answer: Environmental variation

Temporal environmental variation promotes diversity

Assumption: species demographic parameters in current conditions are independentof conditions experienced in the previous generation

Figure modified from Levine & Rees 2002

Variable Constant

The presence of maternal effects violates this assumption

good year bad yearseed

Galloway 2005, Sultan et al. 2006, Dyer et al. 2010, Germain et al. 2013


good year bad year

reinforcing numeric responses

seed



good year bad year

counteracting numeric responses

seed



good year bad yearseed

All existing tests have:

1. included ≤ 3 species only

2. controlled for genotypic differences among individuals

Problem: It is not known how common maternal effects are in biological communities

How common are maternal effects, and how do they compare to regular numeric responses?

Study system: Mediterranean annual plants

hyper-diverse ▪ variable rainfall ▪ many annual species

Papaver dubium

Nemophila menziesii

Uropappus lindleyi

Vulpia microstachysSalvia columbariae

Lasthenia californica

Greenhouse experiment

30 species x 2 env’ts (wet and dry) x 7 pots= total 420 pots

each pot population of ~8 individuals

enumerated seed number, mass per seed, and % seed dormancy

gls: x ~ ME*species




gls: x ~ ME*species

Greenhouse experiment

Raised the offspring generation under similar conditions for first three weeks of growth, measured aboveground biomass as performance measure




gls: x ~ ME*species

*P < 0.05, °P < 0.10

Orienting slide

∆ Seed size

∆ Seed number

maternal effect

regular numeric effects

Orienting slide

∆ Seed size

∆ Seed number

*P < 0.05, °P < 0.10

Orienting slide

∆ Seed size

∆ Seed number

* = significant response

*P < 0.05, °P < 0.10

Orienting slide

∆ Seed size

∆ Seed number

*P < 0.05, °P < 0.10

Result 1: seed size and seed number responses occurred in similar frequencies

∆ Seed size

∆ Seed number

six species

five species

*P < 0.05, °P < 0.10

Result 2: seed size but not seed number responses were bidirectional

∆ Seed size

∆ Seed number

*P < 0.05, °P < 0.10

non-sig. ME effect

sig. ME effect

Result 3: no species responded through both seed size and seed number

∆ Seed size

∆ Seed number

*P < 0.05, °P < 0.10

Result 4: seed size responses are phylogenetically convergent

∆ Seed size

K = 0.09, P = 0.057

*P < 0.05, °P < 0.10

Results summary

Maternal effects:

→ occurred as or more frequently than seed number responses

→ exhibited a fuller range of responses than seed number

→were only observed in species that did not respond through seed number

→were phylogenetically convergent

Results summary

Maternal effects:

→ occurred as or more frequently than seed number responses

→ exhibited a fuller range of responses than seed number

→were only observed in species that did not respond through seed number

→were phylogenetically convergent

Take home messageMaternal effects are common, and should be considered by ecologists

Germain, R. M. and B. Gilbert. (2014). Ecology Letters 17: 662-669

Acknowledgements

Jason Weir

The Gilbert LabBoth AlexsSome of my cats

Bruce HallAndrew Petri

Webpage: rgermain.wordpress.comEmail: [email protected]

Chris BlackfordAlanna LealeAlly MushkaRosemary MartinYvonne ChanAlicia HouKevin HawkshawMags Ngo

Seed size-seed number correlations

Germain & Gilbert 2014 Eco Lett

Seed dormancy responses to the maternal environment vary across species


Seed size effects on offspring performance

r = 0.11, P = 0.047


Botany 2014 Boise, ID

Documents