Blog author: Miranda Salsbery
Citations:
Cole, L. C. (1954). The population consequences of life history phenomena. The Quarterly Review of Biology, 29(2), 103-137.
Ng’habi, K., Viana, M., Matthiopoulos, J., Lyimo, I., Killeen, G., & Ferguson, H. M. (2018). Mesocosm experiments reveal the impact of mosquito control measures on malaria vector life history and population dynamics. Scientific reports, 8(1), 13949.
Background on authors:
LaMont Cole was a professor in zoology at Cornell University. He had broad ranging research staring with herpetology, the subject of which he published is first paper at the age of 19. Most of his career was spent blending mathematical methods and models into the field of ecology. In his later work, he focused more on social aspects of ecology, touching on topics such as thermal pollution and pesticides.
Kija Ng’habi works at the Ifakara Health Institute (IHI) in Tanzania. The IHI is a leading research organization in Africa, specializing in developing, testing and validating innovations for health for over 50 years. Ng’habi has published several other papers on mosquito biology and ecology.
The population consequences of life history phenomena
Here are some of the main points I got out of it:
The goal of this paper was really to introduce population modeling to ecology as well incorporate life history traits into said population modeling. Cole lays out several mathematical equations to do this and uses the human population as his primary example. He mainly focuses on understanding and using r, which he terms “the intrinsic rate of natural increase”. Throughout the paper he explores what happens to population growth when various factors of life history change.
Nonoverlapping generation, such as annual plants, are relatively easy to represent mathematically. Organisms with overlapping generation introduce complexity. Cole’s models for these are built off of a “generation law”. He includes life history traits such as age at first and last reproduction, age specific fecundity and survivorship.
One of the major findings of this paper is that a greatly significant life history trait is not the number of offspring produced in a female’s lifetime, as previously thought, but at what age the offspring are produced. The most rapid population growth is attributes to more reproduction early in life.
Cole also investigate the tradeoffs of a species reproduction only once (semelparity) and multiple times (iteroparity). He found that for semelparity species, the gain in intrinsic rate of growth they would gain from becoming iteroparity is the same as simply increasing their litter size by one. He also found that species with long pre-reproductive periods gain more from iteroparity. Additionally, there is a limiting return on increasing litter size and the number of litters produced.
Age structure also affects population growth. Any increase in longevity not accompanied with an increase in age of last reproduce will lead to a lower birth rate. The portions of the population that are pre-reproductive, reproductive, and post-reproductive directly affect the intrinsic rate of growth
Mesocosm experiments reveal the impact of mosquito control measures on malaria vector life history and population dynamics
Main goal
The mail goal of this paper was to use long term mesocosm experiments to evaluate the effect of combining several anti-mosquito methods on the mosquito’s population and life history traits.
Introduction
Vector control is widespread method for control vector carries diseases such as malaria. Long-lasting insecticidal nets (LLINs) are a common method to reduce the spread of malaria and limit the vectors population (mosquito species Anopheles gambiae). LLINs are limited in their usefulness. They rely on behavioral predisposition of the mosquitos to feed predominantly on humans indoors at night and thus do not prevent them form feeding on other mammals outdoors. Thus, a combination of control measure, such as LLINS in concert with insecticidal eave louvers (EL), or treatment of cattle with the endectocide Ivermectin (IM) by prove to be most effective in lowering vector population and fecundity.
Methods:
The authors established nine replicate populations mosquito populations large mesocosm chambers and allowed them to stabilize. One human and one cow host for were provide as food. After establishment, LLINs were introduced into 6 of the 9 populations, with 3 remaining intervention-free to act as controls for 8 weeks. After this, IM was administered to cattle within 2 of the LLIN-chambers, and insecticide treated ELs were installed in the houses of other 2 LLIN-chambers. Then, IM and EL treatments were swapped between chambers for a final 8 weeks. Mosquito surveillance was done by sampling larvae and adult females every 2–4 weeks. The authors analyzed the data using Bayesian state space models (SSM).
Results:
All three intervention treatments (LLINs, LLINs + EL, LLINs + IM) led to reductions in larval and adult mosquito densities compared to the controls. The introduction of LLINs at the beginning altered this population growth trajectory in all chambers where they were in. Introduction of LLINs was estimated to reduce the weekly adult female survival rate by ~91%. Overall, the combination of LLINs followed by IM had the most disruptive effect on vector populations.
Discussion:
LLINs lead to a reduction in adult mosquito survival, with little subsequent impact on the fecundity of survivors. Extending intervention coverage to an alternative host (e.g. IM on cattle) rather than applying additional protective methods to housing (EL) appeared most effective. This study providesinsights into the fundamental basis of malaria vector population dynamics and emphasizes the importance of vector population dynamics, despite its limitation.
My thoughts:
Both papers stress the importance of including life histories in understanding populations. I find it interesting that the idea Cole proposed 65 years is helping us understand and combat a deadly disease today. While Cole focused on human population dynamic, despite a call for more ecological use, I think he would be impressed today to see just how widespread his concept has become. This Cole paper has over 2,000 citations!
Do you think most people include life history traits in their ecological research? Do you?
I found the papers to complement each other quite well. Cole's paper laid the mathematical groundwork for some concepts used in today's population modeling. Although he was very human-focused, he gave a decent amount of detail to some non-human examples and his concepts show a broad applicability to other ecological systems.
ReplyDeleteI enjoyed the Ng'habi paper a lot. I have done previous work in conservation - working to protect threatened species - so it was very interesting to see ideas of population ecology used for the opposite purpose, i.e. working at reducing/eradicating a species population because it is a vector for disease.
One detail I noticed in Ng'habi's experimental setup was the mesocosm's protection from rainfall. As the authors do later address at the very end of their discussion section, mosquitoes are attracted to water, meaning that rainfall levels could have an impact on the population dynamics in a way that has not been captured with this paper. Although they do not make a strong prediction one way or the other on the impact of rain on their results (stating that it "could either attenuate or increase impacts observed here." p. 7), the effect of rainfall leaves open an area of future study that may be of interest.
- Elizabeth
Cole's paper has me intrigued in how trade-offs play a role in the selection for traits that affect population growth, such as iteroparous versus semelparous reproductive strategies and the age at first reproduction. How might resource availability or climate play a role in these "choices"? Like Elizabeth, I was intrigued by the protection from rainfall in the study by Ng'habi et al., especially since mosquitoes need water in order to reproduce. It seems that water availability is a crucial variable to consider in any study on mosquito populations. Perhaps mosquitoes lay fewer eggs when water is scarce (reducing fecundity) but these eggs are larger and healthier. And perhaps mosquitoes lay more, smaller eggs when water is abundant, producing mosquitoes that are weaker and more prone to insecticides. The Ng'habi study shows that there is potential to study how water availability influences life history traits and to determine how we could use such information to manage disease.
ReplyDeleteThe 2 papers tried to understand population changes with different focus. The Cole’s paper focuses on human demography while the Ng’habi’s paper focuses on disease control. The Cole talked a lot about the mathematical models of geometrical growth and the r value, then he discussed how would that fit into different life strategies and life histories, and how could it be used to predict human demography. I was impressed by the comparison between annual and perennial species, and how their population growth differs. I am not very familiar with demography, and one thing I was confused is “litter” and “litter size”, the way he described the size of “discrete units that offspring are produced in batches”. It makes me wonder what are “consecutive units”.
ReplyDeleteFor Ng’habi’s paper, it is interesting to me to study the population change in a mesocosm. Yes the mesocosm excludes other environmental factors, but I think it is because the aim of this paper is to figure out an efficient method for disease vector control, instead of monitoring the population in the real system.
The two papers compliment each other well with the Cole paper clearly laying some groundwork for future population dynamic studies. Cole's focus on human populations was interesting to me, especially since he stressed the use of a mathematical approach to study non-human populations. Reading the paper was honestly an experience. Cole dates himself on occasion (i.e. talking about the importance of younger individuals for military service) and I am certain some parts of the paper would not be included today (i.e. him insisting 12 year old girls can get pregnant. I cannot be the only one who stopped short when reading this part). Nevertheless, the contributions laid by Cole are interesting and seem to be broadly applicable to many population dynamic studies. I am still a bit unsure what "life history" necessarily means in this case. Cole talks a lot about lifestyle and I am a bit curious to see what a "lifestyle" means for different organisms (because apparently it can change). Also, in his framework, he said human females have a b=1/2 because they have a 50/50 chance of producing another female. Why did he not apply this to other organisms? Why only bring this up when discussing human women?
ReplyDeleteOn another note, I liked the Ng'habi paper. My only nit-pick what that the Results section appeared to be set up initially like a Methods section. Again, this may just be me being a bit nit-picky and not knowing how these papers are typically presented (other than the fact it is from Nature and this seems to be the way their articles are set up). I particularly liked reading a biology paper that takes human behavior into account (i.e can't change people's bedtimes, so they have to work around it). It made for a fascinating read.
Cole’s comparison of population growth to the compound interest equation was interesting. I had only associated compound interest with banking, but I can see how the equation could be used to make simple population growth models. The compound interest equation relies on the variable “r” for the intrinsic rate of natural increase. Cole addresses the different “r” values for the human population based on age at first reproduction in Figure 6. (It is a bit horrifying that he starts the age at first reproduction at 12, but he does have a citation for why he uses that age). I think it would be difficult to calculate “r” for the entire human population because of cultural differences between human populations. Countries with different socioeconomic statuses can have very different reproductive rates, which are represented in the unique shapes of the age-structure population pyramids. The value for “r” in Japan would be very different from the value for “r” in Nigeria. Finding an average “r” for the human population would require analyzing the demographic patterns for each country. Although it is difficult, I assume we have the mathematical models in place for calculating human population growth because the US Census Bureau keeps a “real time” counter of world population growth which shows that the world population grows by approximately 2.5 people every second.
ReplyDeleteThe companion paper looked at vector population dynamics. The state-space population model accounted for the life history of the mosquito. I thought it was important that the paper addressed mosquito behavior as well as human behavior. The nets are effective against mosquitoes while humans are underneath them, but it is unreasonable to assume that humans will be indoors and behind the nets during all hours. The variations in human behavior make it necessary to have the additional vector control interventions in the cattle.
I agree with other comments that the papers are paired well, not just because they look at life history, but because of their approach. The first half of Cole’s paper focuses on methods for calculation of the intrinsic rate of growth (r) from fecundity and survival data, which can be used for forward projection of population size. Ng’habi et al. use state space models to solve the inverse problem, i.e., given these observed population trajectories, how did each treatment impact mosquito fecundity and survival? It’s nice to see how improved computational methods have allowed to find answers to ecological questions that would have intractable in the past.
ReplyDelete- David
Cole is interested in mathematically addressing the aspects of organism life histories that affect population growth and population dynamics. I think much of what he discusses and attempts to model throughout the paper relates to the fundamental concept of interactions between an organism’s genetics (and gene diversity) and a given environment (gene X environment), using humans as a primary example. The variability in reproductive potential, longevity, sex ratio, and consequently the population density that he discusses are all strongly selected upon by an individual’s genes but can be plastic given ephemeral access to resources and changes in environmental conditions (or, “environmental resistance”) that can make certain life histories more/less adaptive. Since I have only worked with these sorts of questions in regards to bark beetles, I related a lot of his models/discussion to them (rather than people), as I have done a lot of thinking about individual/population-level tradeoffs in bark beetles that can latently trigger outbreaks (rapid population growth) as a result of increases in food resources or sharp decreases when a resource disappears. I was intrigued by the equations he outlines for parameters that influence population growth, but I still felt that his limited (lacking) discussion/inclusion in his models of other controls on life history/adaptive change (i.e., parasitism, predation) made them somewhat oversimplified (and humans are even more complex)!
ReplyDeleteThe Ng’habi et al. paper also looked at organism and population-level life history but came from a less theoretical and more management-driven approach, i.e., ways that humans can alter life history trajectories of mosquitoes to control their population levels. However, the state-space model they used to characterize changes in the mosquito life history traits was still essential to determine the effects of the treatments [long-lasting insecticidal nets (LLINs), eave louvers (EL), and ivermectin (EM)]. I thought the paper was an excellent attempt at using model-driven results to inform population management of a serious disease vector. However (as in some of the models the Cole paper) I felt that their binomial models for weekly survival (both larvae and adult females) were limited in that they only accounted for density-dependence and treatment, and none of the environmental factors that drive population density. Although their primary question was related to life history, ignoring environmental variability (particularly in the limited wet habitats mosquitoes need for reproduction) seems dubious given the influence it should have on mosquito population dynamics.
I thought the Cole paper was a good refresher on the history of population growth models. Sometimes I forget that the theory Malthusian catastrophe was contentious..
ReplyDeleteI would say that nowadays population ecologists do generally incorporate life history traits into population growth models. Luckily, the field has progressed since Cole (1954) and ecologists now understand that population growth is a multifaceted process and use approaches that try to capture that complexity, as with stochastic-dynamic simulations and individual-based models that try to scale up individual processes to the population level.
It is interesting to see what has stood the test of time. We saw that some of the concepts Cole brought forth remain relevant, although I thought the companion paper brought forth an interesting point, which was that not all life history traits are easily measured in all populations. Even with modern modeling techniques, it is still not always possible to account for every factor you think might affect your system.