ABSTRACT
Evolutionary rescue occurs when the genetic evolution of adaptation saves a population from decline or extinction after environmental change. The evolution of resistance to pesticides is a special scenario of abrupt environmental change, where rescue occurs under (very) strong selection for one or a few de novo resistance mutations of large effect. Here, a population genetic model of evolutionary rescue with density-dependent population change is developed, with a focus on deriving results that are important to resistance management. Massive stochastic simulations are used to generate observations, which are accurately predicted using analytical approximations. Key results include the probability density function for the time to resistance and the probability of population extinction. The distribution of resistance times shows a lag period, a narrow peak and a long tail. Surprisingly, the mean time to resistance can increase with the strength of selection because, if a mutation does not occur early on, then its emergence is delayed by the pesticide reducing the population size. The probability of population extinction shows a sharp transition, in that when extinction is possible, it is also highly likely. Consequently, population suppression and (local) eradication can be theoretically achievable goals, as novel strategies to delay resistance evolution.
Subject(s)
Pesticides , Biological Evolution , Drug Resistance/genetics , Models, Genetic , Mutation , Selection, Genetic , Animals , Evolution, MolecularABSTRACT
Rotations have been the cornerstone of insecticide resistance management for many decades. In recent years, there has been a resurgence of interest in the use of insecticide mixtures, particularly based on new theoretical models. Here, we present a perspective on the value of rotations to insecticide resistance management, focusing on the interpretation of influential theoretical models. The principles of resistance management have previously been reduced to moderation, saturation and multiple attack. Alongside mixtures and mosaics, rotations have been presented as a strategy of multiple attack in using more than one insecticide. Three explanations have been offered for how rotations delay resistance evolution: counterselection from resistance cost, the relaxation of selection and intergenerational redundant kill. We show that all three explanations can make sense of the comparison of rotations with another resistance-management strategy but have failed to elucidate the principle at work. Overall, we argue that rotations work by moderation, delaying resistance to insecticides through the use of each insecticide less over time. We suggest that the principles of resistance management are recast as moderation, saturation and redundancy. When rotations and mixtures are not conceptualised as competing methods of multiple attack, these strategies can more obviously work together through the complementary principles of moderation and redundancy. Whether solo products or a mixture of products are used, rotations are an effective method of risk management, preserving the arsenal of all effective insecticides for longer. A successful resistance-management plan should make appropriate use of all the principles of resistance management. © 2024 Society of Chemical Industry.
Subject(s)
Insecticides , Insecticides/pharmacology , Insecticide Resistance , Models, TheoreticalABSTRACT
BACKGROUND: Up until the present, pyrethroid-treated bed nets have been a key tool for vector control in the fight against malaria. A global system that sets standards and facilitates procurement has successfully driven down the price of these bed nets to enable more of them to be distributed. As a result of their mass rollout, malaria cases have been significantly reduced, but pyrethroid resistance is now widespread. Going forward, new insecticides have been and continue to be developed for use on bed nets, but it is unclear how to best deploy them for maximum impact. METHODS: Here, an app for the optimization of bed nets based on their insecticide loading concentration and deployment lifespan is presented. Underlying the app are simple models that incorporate the chemical and physical properties of bed nets, and the genetic and ecological properties of resistance evolution in mosquitoes. Where possible, default parameter values are fitted from experimental data. The app numerically searches across a massive number of these simple models with variable loading and lifespan to find their optima under different criteria that constrain the options for vector control. RESULTS: The app is not intended to provide a definite answer about the best bed net design, but allows for the quantative exploration of trade-offs and constraints under different conditions. Here, results for the deployment of a new insecticide are explored under default parameter values across public health budgets for the purchase of bed nets. Optimization can lead to substantial gains in the average control of the mosquito population, and these gains are comparatively greater with lower budgets. Whilst optimizing a bed net within the constraints of the incentives of the existing system of standards and procurement leads to substantially greater control than not optimizing the bed net, optimizing the bed net without constraints leads to yet substantially greater control. The most important factor in this optimization is coverage, which depends on the price per bed net. With this in mind, it is unsurprising that the optimization for plausible budgets suggests that a pyrethroid would be the preferred partner for a new insecticide under current constraints because it is cost-effective in the balance of being less expensive than the new insecticide but also less effective due to pre-existing resistance. Surprisingly, a pyrethroid is shown to be an effective partner for a new insecticide in this model because of its contribution to resistance management in delaying the onset of resistance to the new insecticide. CONCLUSIONS: This study highlights the importance of trade-offs in the design of bed nets for vector control. Further, it suggests that there are challenges in the roll-out of bed nets with new insecticides because of the constraints imposed by the global system of standards and procurement, which currently fails to adequately incentivize important considerations in bed net design like resistance management.
Subject(s)
Anopheles , Insecticide-Treated Bednets , Insecticides , Malaria , Mobile Applications , Pyrethrins , Animals , Humans , Insecticides/pharmacology , Mosquito Control/methods , Insecticide Resistance , Mosquito Vectors , Pyrethrins/pharmacology , Malaria/epidemiologyABSTRACT
The use of insecticide mixtures for resistance management has been a controversial topic for many decades. Here, we provide a reassessment of the fundamental theory of insecticide mixtures. First, we examine how mixtures differ from other strategies. We suggest that the fundamental strategy concept of a mixture is defined by the simultaneous use of insecticides and their overlapping exposure. Second, we provide a simple, illustrative model to show how mixtures affect resistance evolution. Following the existing literature, we identify a role for 'redundant kill' acting against resistant individuals, which we link to the overlapping exposure of insecticides. We also identify the occurrence of 'additional kill' acting against susceptible individuals, which is the immediate consequence of the simultaneous use of insecticides. Third, we take a basic approach to the comparison of mixtures and other strategies using a simple model. We find that a common comparison of the time to resistance alone leaves the effects of additional kill unaccounted for. Moreover, we demonstrate that different approaches to comparison can lead to different results because of biases that are introduced in the comparison setup. Fourth, still using the same model, we showcase a more sophisticated approach to comparison using optimised strategies. We find that optimised mixtures always perform better than other strategies due to the combination of redundant and additional kill. We suggest that the comparison of optimised strategies is unbiased because each strategy is performing the best that it can. On this basis, in theory (but not necessarily practice), we believe that mixtures are better than other strategies and, through the steps of our argument, we can tie this success back to the fundamental properties (of simultaneous use and overlapping exposure) that distinguish mixtures from other strategy concepts. © 2022 The Authors. Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.
Subject(s)
Insecticides , Humans , Insecticides/pharmacology , Insecticide Resistance , Insect Control/methodsABSTRACT
BACKGROUND: The program to eradicate malaria is at a critical juncture as a new wave of insecticides for mosquito control enter their final stages of development. Previous insecticides have been deployed one-at-a-time until their utility was compromised, without the strategic management of resistance. Recent investment has led to the near-synchronous development of new insecticides, and with it the current opportunity to build resistance management into mosquito-control methods to maximize the chance of eradicating malaria. METHODS: Here, building on the parameter framework of an existing mathematical model, resistance-management strategies using multiple insecticides are compared to suggest how to deploy combinations of available and new insecticides on bed nets to achieve maximum impact. RESULTS: Although results support the use of different strategies in different settings, deploying new insecticides ideally together in (or at least as a part of) a mixture is shown to be a robust strategy across most settings. CONCLUSIONS: Substantially building on previous works, alternative solutions for the resistance management of new insecticides to be used in bed nets for malaria vector control are found. The results support a mixture product concept as the most robust way to deploy new insecticides, even if they are mixed with a pyrethroid that has lower effectiveness due to pre-existing resistance. This can help deciding on deployment strategies and policies around the sustainable use of these new anti-malaria tools.
Subject(s)
Anopheles , Insecticide-Treated Bednets , Malaria , Animals , Insecticide Resistance , Malaria/prevention & control , Mosquito VectorsABSTRACT
The use of multiple pesticides or drugs can lead to a simultaneous selection pressure for resistance alleles at different loci. Models of resistance evolution focus on how this can delay the spread of resistance through a population, but often neglect how this can also reduce the probability that a resistance allele spreads. This neglected factor has been studied in a parallel literature as selective interference. Models of interference use alternative constructions of fitness, where selection coefficients from different loci either add or multiply. Although these are equivalent under weak selection, the two constructions make alternative predictions under the strong selection that characterizes resistance evolution. Here, simulations are used to examine the effects of interference on the probability of fixation and time to fixation of a new and strongly beneficial mutation in the presence of another strongly beneficial allele with variable starting frequency. The results from simulations show a complicated pattern of effects. The key result is that, under multiplicativity, the presence of the strongly beneficial allele leads to a small reduction in the probability of fixation for the new beneficial mutation up to ~10%, and a negligible increase in the average time to fixation up to ~2%, whereas under additivity, the effect is more substantial at up to ~50% for the probability of fixation and ~100% for the average time to fixation. Consequently, the effect of interference is only an important feature of resistance evolution under additivity. Current evidence from studies of experimental evolution provides widespread support for the basic features of additivity, which suggests that interference may afford resistance a different pattern of evolution than other adaptations: rather than the gradual and simultaneous selection of many alleles with small effects, the rapid evolution of resistance may involve the sequential selection of alleles with large effects.
Subject(s)
Models, Genetic , Selection, Genetic , Alleles , Evolution, Molecular , Genetics, Population , Mutation , ProbabilityABSTRACT
Although it is generally accepted that geography is a major factor shaping human genetic differentiation, it is still disputed how much of this differentiation is a result of a simple process of isolation-by-distance, and if there are factors generating distinct clusters of genetic similarity. We address this question using a geographically explicit simulation framework coupled with an Approximate Bayesian Computation approach. Based on six simple summary statistics only, we estimated the most probable demographic parameters that shaped modern human evolution under an isolation by distance scenario, and found these were the following: an initial population in East Africa spread and grew from 4000 individuals to 5.7 million in about 132 000 years. Subsequent simulations with these estimates followed by cluster analyses produced results nearly identical to those obtained in real data. Thus, a simple diffusion model from East Africa explains a large portion of the genetic diversity patterns observed in modern humans. We argue that a model of isolation by distance along the continental landmasses might be the relevant null model to use when investigating selective effects in humans and probably many other species.
Subject(s)
Geography , Models, Genetic , Demography , Genetics, Population , HumansABSTRACT
Gradients of variation--or clines--have always intrigued biologists. Classically, they have been interpreted as the outcomes of antagonistic interactions between selection and gene flow. Alternatively, clines may also establish neutrally with isolation by distance (IBD) or secondary contact between previously isolated populations. The relative importance of natural selection and these two neutral processes in the establishment of clinal variation can be tested by comparing genetic differentiation at neutral genetic markers and at the studied trait. A third neutral process, surfing of a newly arisen mutation during the colonization of a new habitat, is more difficult to test. Here, we designed a spatially explicit approximate Bayesian computation (ABC) simulation framework to evaluate whether the strong cline in the genetically based reddish coloration observed in the European barn owl (Tyto alba) arose as a by-product of a range expansion or whether selection has to be invoked to explain this colour cline, for which we have previously ruled out the actions of IBD or secondary contact. Using ABC simulations and genetic data on 390 individuals from 20 locations genotyped at 22 microsatellites loci, we first determined how barn owls colonized Europe after the last glaciation. Using these results in new simulations on the evolution of the colour phenotype, and assuming various genetic architectures for the colour trait, we demonstrate that the observed colour cline cannot be due to the surfing of a neutral mutation. Taking advantage of spatially explicit ABC, which proved to be a powerful method to disentangle the respective roles of selection and drift in range expansions, we conclude that the formation of the colour cline observed in the barn owl must be due to natural selection.
Subject(s)
Biological Evolution , Genetics, Population , Pigmentation/genetics , Selection, Genetic , Strigiformes/genetics , Animals , Bayes Theorem , Computer Simulation , Europe , Microsatellite Repeats , Models, Biological , Sequence Analysis, DNAABSTRACT
A species of Galápagos tortoise endemic to Española Island was reduced to just 12 females and three males that have been bred in captivity since 1971 and have produced over 1700 offspring now repatriated to the island. Our molecular genetic analyses of juveniles repatriated to and surviving on the island indicate that none of the tortoises sampled in 1994 had hatched on the island versus 3% in 2004 and 24% in 2007, which demonstrates substantial and increasing reproduction in situ once again. This recovery occurred despite the parental population having an estimated effective population size <8 due to a combination of unequal reproductive success of the breeders and nonrandom mating in captivity. These results provide guidelines for adapting breeding regimes in the parental captive population and decreasing inbreeding in the repatriated population. Using simple morphological data scored on the sampled animals, we also show that a strongly heterogeneous distribution of tortoise sizes on Española Island observed today is due to a large variance in the number of animals included in yearly repatriation events performed in the last 40 years. Our study reveals that, at least in the short run, some endangered species can recover dramatically despite a lack of genetic variation and irregular repatriation efforts.
ABSTRACT
The Americas were the last continents to be populated by humans, and their colonization represents a very interesting chapter in our species' evolution in which important issues are still contentious or largely unknown. One difficult topic concerns the details of the early peopling of Beringia, such as for how long it was colonized before people moved into the Americas and the demography of this occupation. A recent work using mitochondrial genome (mtDNA) data presented evidence for a so called "three-stage model" consisting of a very early expansion into Beringia followed by approximately 20,000 years of population stability before the final entry into the Americas. However, these results are in disagreement with other recent studies using similar data and methods. Here, we reanalyze their data to check the robustness of this model and test the ability of Native American mtDNA to discriminate details of the early colonization of Beringia. We apply the Bayesian Skyline Plot approach to recover the past demographic dynamic underpinning these events using different mtDNA data sets. Our results refute the specific details of the "three-stage model", since the early stage of expansion into Beringia followed by a long period of stasis could not be reproduced in any mtDNA data set cleaned from non-Native American haplotypes. Nevertheless, they are consistent with a moderate population bottleneck in Beringia associated with the Last Glacial Maximum followed by a strong population growth around 18,000 years ago as suggested by other recent studies. We suggest that this bottleneck erased the signals of ancient demographic history from recent Native American mtDNA pool, and conclude that the proposed early expansion and occupation of Beringia is an artifact caused by the misincorporation of non-Native American haplotypes.
Subject(s)
DNA, Mitochondrial/genetics , Indians, North American/genetics , Algorithms , Bayes Theorem , Biological Evolution , Brazil , Emigration and Immigration , Evolution, Molecular , Genetic Variation , Genetics, Population , Genome, Human , Haplotypes , Humans , Models, Statistical , Time FactorsABSTRACT
It is well accepted that the Americas were the last continents reached by modern humans, most likely through Beringia. However, the precise time and mode of the colonization of the New World remain hotly disputed issues. Native American populations exhibit almost exclusively five mitochondrial DNA (mtDNA) haplogroups (A-D and X). Haplogroups A-D are also frequent in Asia, suggesting a northeastern Asian origin of these lineages. However, the differential pattern of distribution and frequency of haplogroup X led some to suggest that it may represent an independent migration to the Americas. Here we show, by using 86 complete mitochondrial genomes, that all Native American haplogroups, including haplogroup X, were part of a single founding population, thereby refuting multiple-migration models. A detailed demographic history of the mtDNA sequences estimated with a Bayesian coalescent method indicates a complex model for the peopling of the Americas, in which the initial differentiation from Asian populations ended with a moderate bottleneck in Beringia during the last glacial maximum (LGM), around approximately 23,000 to approximately 19,000 years ago. Toward the end of the LGM, a strong population expansion started approximately 18,000 and finished approximately 15,000 years ago. These results support a pre-Clovis occupation of the New World, suggesting a rapid settlement of the continent along a Pacific coastal route.