Descrizione del progetto
Sondare le differenze fenotipiche nelle libellule della medesima specie
Un centinaio di anni di ricerca ecologica ed evoluzionistica ci ha permesso di migliorare la nostra comprensione su fenotipi alternativi e polimorfismi cromatici in natura. I processi evoluzionistici come la selezione dipendente dalla frequenza negativa potrebbero aiutare a spiegare cosa mantiene la diversità genetica all’interno di una popolazione. Il progetto PhenoDim, finanziato dall’UE, intende condurre analisi fototipiche esaustive sul dimorfismo sessuale e sul polimorfismo cromatico limitato alle femmine nelle libellule, attraverso l’uso di una fenotipizzazione ad alta produttività. L’obiettivo generale è quello di capire se una libellula femmina che ha il colore del maschio varia anche in altri aspetti, potenzialmente connessi ai tratti fenotipici come la morfologia, il comportamento e lo sviluppo.
Obiettivo
During over 100 years of ecological and evolutionary research, the study of colour polymorphisms has helped to improve our understanding of phenotypic diversification in nature. Polymorphisms are typically expected to be maintained within local populations by evolutionary processes, such as negative frequency-dependent selection. Yet there are only few studies that have quantified multiple axes of correlated phenotypic variation in addition to colour, which limits our current understanding of the emergence and evolutionary maintenance of genetic colour polymorphisms, for example, via correlational selection, linkage or pleiotropic effects. To address this long standing issue, I here propose a series of research projects that use computer-driven high throughput phenotyping to conduct comprehensive phenotypic analyses (phenomics) of a sexual dimorphism and female-limited colour polymorphisms in damselflies (Odonata; Zygoptera). The overarching goal of these projects is to investigate if and how much female damselfly colour morphs that mimic male colouration to increase their fitness (androchrome females) also vary in other, potentially correlated phenotypic traits, such as morphology, behavior and development. First, I will analyze an existing image data set of evolutionary time series data of Swedish damselfly populations, using a combination of classic computer vision and deep learning algorithms to quantify axes of phenotypic variation that have not yet been investigated. Second, I will conduct a mesocosm experiment using individuals from the same local Swedish damselfly populations to investigate whether androchrome females show alternative developmental trajectories. Third, I will conduct in-depth phenotyping of an existing image dataset of different damselfly species to investigate the prevalence of androchromatism and male mimicry in other traits across a large damselfly phylogeny.
Campo scientifico
- natural sciencesbiological sciencesbiological morphologycomparative morphology
- engineering and technologymaterials engineeringcolors
- natural sciencescomputer and information sciencesartificial intelligencecomputer vision
- natural sciencesbiological sciencesevolutionary biology
- natural sciencescomputer and information sciencesartificial intelligencemachine learningdeep learning
Parole chiave
Programma(i)
Argomento(i)
Meccanismo di finanziamento
MSCA-IF - Marie Skłodowska-Curie Individual Fellowships (IF)Coordinatore
22100 Lund
Svezia