The portfolio effect cushions mosquito populations and malaria transmission against vector control interventions
Portfolio effects were first described as a basis for mitigating against financial risk by diversifying investments. Distributing investment across several different assets can stabilize returns and reduce risks by statistical averaging of individual asset dynamics that often correlate weakly or negatively with each other. The same simple probability theory is equally applicable to complex ecosystems, in which biological and environmental diversity stabilizes ecosystems against natural and human-mediated perturbations. Given the fundamental limitations to how well the full complexity of ecosystem dynamics can be understood or anticipated, the portfolio effect concept provides a simple framework for more critical data interpretation and pro-active conservation management. Applied to conservation ecology purposes, the portfolio effect concept informs management strategies emphasizing identification and maintenance of key ecological processes that generate complexity, diversity and resilience against inevitable, often unpredictable perturbations.
Applied to the reciprocal goal of eliminating the least valued elements of global biodiversity, specifically lethal malaria parasites and their vector mosquitoes, simply understanding the portfolio effect concept informs more cautious interpretation of surveillance data and simulation model predictions. Malaria transmission mediated by guilds of multiple vectors in complex landscapes, with highly variable climatic and meteorological conditions, as well as changing patterns of land use and other human behaviours, will systematically tend to be more resilient to attack with vector control than it appears based on even the highest quality surveillance data or predictive models.
Malaria vector control programmes may need to be more ambitious, interpret their short-to-medium term assessments of intervention impact more cautiously, and manage stakeholder expectations more conservatively than has often been the case thus far.
KeywordsMalaria Plasmodium Anopheles Mosquito Vector control Elimination Ecology
long lasting insecticidal nets
indoor residual spraying
Conservation biologists have recently adopted the portfolio effect concept from economics , to guide their thinking in relation to ecosystem conservation . The implications of such simple probability theory for financial investments are rather obvious and now widely accepted: diversification stabilizes investment portfolios, thereby reducing risks of catastrophic losses . Distributing investment across several different assets can stabilize returns and reduce risks by statistical averaging of individual asset dynamics that often correlate weakly or negatively with each other .
The same simple probability theory  is equally applicable to complex ecosystems, which are buffered against natural and human-mediated perturbations by biological and environmental diversity . Rather than rely on prescriptive model predictions, the uncertainties of which are determined by fundamental limitations to how well the full complexity of ecosystem dynamics can be understood or anticipated, the portfolio effect concept provides a simple framework for more critical data interpretation and pro-active conservation management . Merely understanding the portfolio effect concept informs management strategies emphasizing identification and maintenance of key ecological processes that generate complexity, diversity and resilience against inevitable and often unpredictable perturbations .
Implications for malaria vector control and surveillance
The implications of the portfolio effect concept should also be considered when interpreting malaria vector surveillance data and the predictions of simulation models fitted to them. When considering how uncertain models might be, it is important to distinguish between the likely causes of unbiased imprecision and systematic inaccuracy. The portfolio effect introduces the latter: by design, mathematical models are deliberately less complex than the biological system they are intended to mimic [3, 4, 5], and no dataset can capture all the different circumstances a real biological system experiences. There is, therefore, an inevitable tendency for mathematical models to underestimate the complexity and associated resilience of natural biological systems. Expressed in simple interpretational terms, mosquito populations and malaria transmission will tend be more resilient against control efforts than face-value interpretation of data or predictive mathematical models suggest.
The ubiquitous and extreme heterogeneities of vector density and vectorial capacity that occur across remarkably fine geographic scales have long been recognized as crucial factors underpinning the notorious intransigence of malaria transmission to intervention efforts [6, 7, 8]. However, beyond heterogeneities of vector density resulting in local foci where transmission is far more intense and stable than the landscape-wide average, vector biodiversity and heterogeneities in the environments they live in create portfolio effects that diversify the properties of malaria transmission.
Modelling analyses that incorporated heterogeneities of mosquito behaviour were centrally important to the illustrations of how residual malaria transmission [9, 10, 11, 12] persists so robustly in Africa after scale-up of indoor residual spraying (IRS) [13, 14]. Since then, a variety of models have been used to illustrate this same point [15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]. However, no existing model captures the full range of all relevant mosquito behaviour in real transmission systems with biodiversity spanning dozens of vectors , several of which may occur in any given setting. While such models can be improved, progress towards more realistic representations of complex real-life vector systems will be limited by data and understanding for the foreseeable future [5, 32]. In the meantime, it may be prudent to bear in mind the following rule of thumb: the more diverse and variable the life histories of malaria vectors are, the less likely it is that any given vector control approach with eliminate all the malaria transmission they mediate.
For example, the more mosquito species a malaria parasite can use as a vector, the higher the probability that one or more of those species will become resistant to any given insecticide compared to a situation where only a single vector species is involved. Given that there is a stochastic element to resistance evolution, the more vector species are present, the more likely that at least one of them will become physiologically resistant to insecticides and continue to mediate transmission despite high coverage of long-lasting insecticidal nets (LLINs) and/or IRS. For example, while Anopheles gambiae has been greatly reduced in numbers across many parts of Africa following scale-up of pyrethroid-based LLINs [33, 34], highly pyrethroid-resistant Anopheles funestus  may persist and mediate intense transmission .
Also, differences in the behaviour of mosquito species have long been known to render malaria transmission frustratingly resilient against attack with IRS [14, 37, 38]. Behaviourally selective vector control interventions, such as LLINs and IRS, have successfully eliminated entire populations of some of the world’s most important malaria vectors, such as An. gambiae and An. funestus in Africa, Anopheles punctulatus and Anopheles koliensis in Oceania, or Anopheles darlingi in South America. However, elimination of malaria transmission remains elusive in most settings because mosquito species persist which are less efficient vectors but also exhibit outdoor resting and feeding behaviours that are far less vulnerable to these indoor-targeted approaches [9, 10, 28, 34, 39].
Furthermore, fine-scale environmental variations in the relative abundance and availabilities of essential blood host and resting site resources can drive huge variations in the behavioural choices that mosquitoes exhibit in different parts of a given landscape. Taking Anopheles arabiensis as an African example of an important vector of residual malaria transmission that exhibits notoriously plastic feeding behaviours, the proportion of indoor-feeding mosquitoes that rest indoors until the following morning can vary by two orders of magnitude . More tellingly, An. arabiensis can exhibit both extremes of feeding predominantly on either people or cattle, even in different family compounds within the same small village [41, 42].
The diversity of behaviours expressed by a single vector species through phenotypic plasticity is further exacerbated by the fact that they usually co-exist alongside other vectors with different behavioural preferences (Fig. 1). Such guilds of multiple vectors may span a remarkably wide range of behavioural phenotypes, and the African scenario presented in Fig. 1 is far less biodiverse than many settings in southeast Asia [9, 45, 46, 47]. It will therefore be necessary to design packages of complementary vector control interventions based on the range of behaviours observed in nationally representative surveys, rather than their mean values, so that these intervention combinations are broadly applicable and robust to local variations in the behaviours targeted by each component control measure . Such a multi-intervention approach would also help address the urgent need to implement insecticide resistance management strategies , by exploiting multiple interventions that allow different, complementary insecticide classes to be deployed as combinations delivered through distinct products.
All these examples of the complexities that bolster malaria transmission against vector control interventions can be bamboozling and distract from very simple common principles that underlie them all. Malaria parasite populations that typically spread their reproductive bets across two or more vectors with different behaviours, ecological niches or seasonality dynamics will systematically be more difficult to eliminate than in the rare settings with a single vector species. Furthermore, where individual vector species spread their own reproductive bets across multiple aquatic habitat types, resting sites or blood sources, this creates refugia that limit the impact of any given vector control measure applied in any given time and place. And no matter how much detail we try to capture in mathematical models of vector biology and malaria transmission, they will always under-represent the full complexity and diversity of those interactions, so they are biased towards underestimating the resilience of malaria transmission against vector control. Whatever the shape of the expected response curve following introduction of a new vector control measure, the portfolio effect will tend to flatten it out to some extent. Given that the magnitude of such portfolio effects are unknown in any given location, the only sensible way to deal with their implications is to emphasize the need for cautious interpretation of entomological surveillance data, as well as simulation models extrapolating these trends into the future.
GFK drafted the manuscript in consultation with TR. Both authors read and approved the final manuscript.
We kindly thank Ms. Eleanor Campos Killeen for drawing Fig. 3.
The authors declare that they have no competing interests.
Availability of data and materials
Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.
Consent for publication
Ethics approval and consent to participate
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 1.Markowitz H. Portfolio selection. J Finance. 1952;7:77–91.Google Scholar
- 9.Durnez L, Coosemans M. Residual transmission of malaria: an old issue for new approaches. In: Manguin S, editor. Anopheles mosquitoes—new insights into malaria vectors. Rijeka: Intech; 2013. p. 671–704.Google Scholar
- 11.WHO. Guidance note-control of residual malaria parasite transmission. Geneva: World Health Organization Global Malaria Programme; 2014. p. 5.Google Scholar
- 12.WHO. Malaria terminology. WHO/HTM/GMP/2016. Geneva: World Health Organization; 2016. p. 31.Google Scholar
- 14.Molineaux L, Gramiccia G. The Garki project. Geneva: World Health Organ; 1980.Google Scholar
- 28.Killeen GF, Kiware SS, Okumu FO, Sinka ME, Moyes CL, Massey NC, et al. Going beyond personal protection against mosquito bites to eliminate malaria transmission: population suppression of malaria vectors that exploit both human and animal blood. BMJ Glob Health. 2017;2:e000198.CrossRefPubMedPubMedCentralGoogle Scholar
- 32.Killeen GF, Chaki PP, Reed TE, Moyes CL, Govella NJ. Entomological surveillance as a cornerstone of malaria elimination: a critical appraisal. In: Dev V, Manguin S, editors. Towards malaria elimination—a leap forward. London: InTech; 2018. p. 403–29.Google Scholar
- 35.Riveron JM, Ibrahim SS, Mulamba C, Djouaka R, Irving H, Wondji MJ, et al. Genome-wide transcription and functional analyses reveal heterogeneous molecular mechanisms driving pyrethroids resistance in the major malaria vector Anopheles funestus across Africa. G3 (Bethesda). 2017;7:1819–32.Google Scholar
- 36.Kaindoa EW, Matowo NS, Ngowo HS, Mkandawile G, Mmbando A, Finda M, et al. Interventions that effectively target Anopheles funestus mosquitoes could significantly improve control of persistent malaria transmission in south-eastern Tanzania. PLoS One. 2017;12:e0177807.CrossRefPubMedPubMedCentralGoogle Scholar
- 40.Killeen GF, Kiware SS, Seyoum A, Gimnig JE, Corliss GF, Stevenson J, et al. Comparative assessment of diverse strategies for malaria vector population control based on measured rates at which mosquitoes utilize targeted resource subsets. Malar J. 2014;13:338.CrossRefPubMedPubMedCentralGoogle Scholar
- 44.Silver JB. Blood feeding and its epidemiological significance. Mosquito ecology: field sampling methods. Dordrecht: Springer; 2008. p. 677–769.Google Scholar
- 48.WHO. Global plan for insecticide resistance management in malaria vectors (GPIRM). Geneva: World Health Organization; 2012. p. 130.Google Scholar
- 49.Muirhead-Thomson RC. Mosquito behaviour in relation to malaria transmission and control in the tropics. London: Edward Arnold & Co.; 1951.Google Scholar
- 50.Holstein MH. Biology of Anopheles gambiae. Geneva: World Health Organization; 1954.Google Scholar
- 51.Gillies MT, De Meillon B. The Anophelinae of Africa South of the Sahara (Ethiopian zoogeographical region). Johannesburg: South African Institute for Medical Research; 1968.Google Scholar
- 57.Briercliffe R, Dalryimple-Champney W. Discussion on the malaria epidemic in Ceylon 1934–1935. Proc R Soc Med. 1935;29:537–62.Google Scholar
- 63.MacDonald G. The epidemiology and control of malaria. London: Oxford University Press; 1957.Google Scholar
- 73.Gillies MT, Coetzee M. A supplement to the Anophelinae of Africa South of the Sahara (Afrotropical region). Johannesburg: South Afr Med Res Institute; 1987.Google Scholar
- 74.Kawada H, Dida GO, Sonye G, Njenga SM, Mwandawiro C, Minakawa N. Reconsideration of Anopheles rivulorum as a vector of Plasmodium falciparum in western Kenya: some evidence from biting time, blood preference, sporozoite positive rate, and pyrethroid resistance. Parasit Vectors. 2012;5:230.CrossRefPubMedPubMedCentralGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.