One of the most intriguing questions in evolution is how organisms exhibit suitable phenotypic variation to rapidly adapt in novel selective environments. Such variability is crucial for evolvability, but poorly understood. In particular, how can natural selection favour developmental organisations that facilitate adaptive evolution in previously unseen environments? Such a capacity suggests foresight that is incompatible with the short-sighted concept of natural selection. A potential resolution is provided by the idea that evolution may discover and exploit information not only about the particular phenotypes selected in the past, but their underlying structural regularities: new phenotypes, with the same underlying regularities, but novel particulars, may then be useful in new environments. If true, we still need to understand the conditions in which natural selection will discover such deep regularities rather than exploiting ‘quick fixes’ (i.e., fixes that provide adaptive phenotypes in the short term, but limit future evolvability). Here we argue that the ability of evolution to discover such regularities is formally analogous to learning principles, familiar in humans and machines, that enable generalisation from past experience. Conversely, natural selection that fails to enhance evolvability is directly analogous to the learning problem of over-fitting and the subsequent failure to generalise. We support the conclusion that evolving systems and learning systems are different instantiations of the same algorithmic principles by showing that existing results from the learning domain can be transferred to the evolution domain. Specifically, we show that conditions that alleviate over-fitting in learning systems successfully predict which biological conditions (e.g., environmental variation, regularity, noise or a pressure for developmental simplicity) enhance evolvability. This equivalence provides access to a well-developed theoretical framework from learning theory that enables a characterisation of the general conditions for the evolution of evolvability.

Adaptive plasticity allows organisms to cope with environmental change, thereby increasing the population’s long-term fitness. However, individual selection can only compare the fitness of individuals within each generation: if the environment changes more slowly than the generation time (i.e., a coarse-grained environment) a population will not experience selection for plasticity even if it is adaptive in the long-term. How does adaptive plasticity then evolve? One explanation is that, if competing alleles conferring different degrees of plasticity persist across multiple environments, natural selection between genetic lineages could select for adaptive plasticity (lineage selection). We show that adaptive plasticity can evolve even in the absence of such lineage selection. Instead, we propose that adaptive plasticity in coarse-grained environments evolves as a by-product of inefficient short-term natural selection: populations that rapidly evolve their phenotypes in response to selective pressures follow short-term optima, with the result that they have reduced long-term fitness across environments. Conversely, populations that accumulate limited genetic change within each environment evolve long-term adaptive plasticity even when plasticity incurs short-term costs. These results remain qualitatively similar regardless of whether we decrease the efficiency of natural selection by increasing the rate of environmental change or decreasing mutation rate, demonstrating that both factors act via the same mechanism. We demonstrate how this mechanism can be understood through the concept of learning rate. Our work shows how plastic responses that are costly in the short term, yet adaptive in the long term, can evolve as a by-product of inefficient short-term selection, without selection for plasticity at either the individual or lineage level.

It has been hypothesized that one of the main reasons evolution has been able to produce such impressive adaptations is because it has improved its own ability to evolve -"the evolution of evolvability". Rupert Riedl, for example, an early pioneer of evolutionary developmental biology, suggested that the evolution of complex adaptations is facilitated by a developmental organization that is itself shaped by past selection to facilitate evolutionary innovation. However, selection for characteristics that enable future innovation seems paradoxical: natural selection cannot favor structures for benefits they have not yet produced; and favoring characteristics for benefits that have already been produced does not constitute future innovation.Here we resolve this paradox by exploiting a formal equivalence between the evolution of evolvability and learning systems. We use the conditions that enable simple learning systems to generalize, i.e., to use past experience to improve performance on previously unseen, future test cases, to demonstrate conditions where natural selection can systematically favor developmental organizations that benefit future evolvability. Using numerical simulations of evolution on highly epistatic fitness landscapes, we illustrate how the structure of evolved gene regulation networks can result in increased evolvability capable of avoiding local fitness peaks and discovering higher fitness phenotypes. Our findings support Riedl's intuition: Developmental organizations that "mimic" the organization of constraints on phenotypes can be favored by short-term selection and also facilitate future innovation. Importantly, the conditions that enable the evolution of such surprising evolvability follow from the same non-mysterious conditions that permit generalization in learning systems.

Adaptive plasticity allows organisms to cope with environmental change, thereby increasing the population's long-term fitness. However, individual selection can only compare the fitness of individuals within each generation: if the environment changes more slowly than the generation time (i.e., a coarse-grained environment) a population will not experience selection for plasticity even if it is adaptive in the long-term. How does adaptive plasticity then evolve? One explanation is that, if competing alleles conferring different degrees of plasticity persist across multiple environments, natural selection between lineages carrying those alleles could select for adaptive plasticity (lineage selection).We show that adaptive plasticity can evolve even in the absence of such lineage selection. Instead, we propose that adaptive plasticity in coarse-grained environments evolves as a by-product of inefficient short-term natural selection. In our simulations, populations that can efficiently respond to selective pressures follow short-term, local, optima and have lower long-term fitness. Conversely, populations that accumulate limited genetic change within each environment evolve long-term adaptive plasticity even when plasticity incurs short-term costs. These results remain qualitatively similar regardless of whether we decrease the efficiency of natural selection by increasing the rate of environmental change or decreasing mutation rate, demonstrating that both factors act via the same mechanism. We demonstrate how this mechanism can be understood through the concept of learning rate.Our work shows how plastic responses that are costly in the short term, yet adaptive in the long term, can evolve as a by-product of inefficient short-term selection, without selection for plasticity at either the individual or lineage level.Organisms that live in variable environments are often subject to opposing selective 2 pressures, either temporal or spatial, such that intermediate generalist phenotypes have 3 decreased fitness across all environments. Rather than evolving a generalist phenotype, 4 populations can keep adapting to each environmental condition as they encounter them, 5 a process known as adaptive tracking [1, 2]. Populations that evolve via adaptive 6 tracking avoid the trade-offs paid by generalist phenotypes, but need time to adapt to 7 each new environment. Consequently, the population experiences reduced fitness after 8 each environmental change. Both populations that evolve a generalist phenotype and 9 those that evolve by adaptive tracking thus have reduced fitness in the long term. By 10 contrast, adaptive phenotypic plasticity allows individuals to maintain an adaptive fit 11 PLOS 1/19 between phenotype and environment: plastic individuals produce only high fitness 12 phenotypes by responding appropriately to environmental cues. Populations evolving 13 adaptive plasticity thus avoid both the fitness loss arising from trade-offs of generalist 14 phenotypes and the fitness loss that tracking populations suffer af...

The paper discusses the outcomes of the conference organized by the InDeal project. The conference took place on 12 December 2018 in Montpellier as part of the EnerGaia energy forum 2018. A holistic interdisciplinary approach for district heating and cooling (DHC) networks is presented that integrates heterogeneous innovative technologies from various scientific sectors. The solution is based on a multi-layer control and modelling framework that has been designed to minimize the total plant production costs and optimize heating/cooling distribution. Artificial intelligence tools are employed to model uncertainties associated with weather and energy demand forecasts, as well as quantify the energy storage capacity. Smart metering devices are utilized to collect information about all the crucial heat substations’ parameters, whereas a web-based platform offers a unique user environment for network operators. Three new technologies have been further developed to improve the efficiency of pipe design of DHC systems: (i) A new sustainable insulation material for reducing heat losses, (ii) a new quick-fit joint for an easy installation, and (iii) a new coating for reducing pressure head losses. The results of a study on the development and optimization of two energy harvesting systems are also provided. The assessment of the environmental, economic and social impact of the proposed holistic approach is performed through a life cycle analysis. The validation methodology of the integrated solution is also described, whereas conclusions and future work are finally given.