Another failure of peer review, due to corrupt temperature data from a single station

UPDATED – see note below. This appeared in Eurekalert a couple of weeks ago. The headline, at first glance, looks like good news, right? “Global warming” is killing off tsetse flies, what’s not to like? Read on, because a photo really is worth a thousand words.

Zambezi Valley may soon be too hot for tsetse flies

From STELLENBOSCH UNIVERSITY via Eurekalert

A new study, based on 27 years of data from Mana Pools National Park in Zimbabwe, suggests that temperature increases over the last three decades have already caused major declines in local populations of tsetse flies.

This analysis, published in the journal PLOS Medicine this week, provides a first step in linking temperature to the risk of sleeping sickness in Africa.

Tsetse are blood-feeding insects that transmit trypanosome pathogens which cause sleeping sickness in humans across sub-Saharan Africa. Without treatment, the disease is fatal. Parasites of this genus also cause nagana, animal African trypanosomiasis (AAT), in livestock. The most recent global estimates indicate that AAT kills approximately one million cattle per year.

The study is based on prolonged laboratory and field measures of fly densities from the 1990s, and nearly continuous records of climatic data since 1975, recorded by researchers based at the Rekomitjie Research Station in the park. Since the 1990s, catches of tsetse flies from cattle in the park declined from more than 50 flies per animal per catching session in 1990, to less than 1 fly per 10 catching sessions in 2017. Since 1975, mean daily temperatures have risen by nearly 1° C and by around 2° C in the hottest month of November.

Researchers from the Liverpool School of Tropical Medicine (LSTM), the South African Centre of Excellence for Epidemiological Modelling and Analysis (SACEMA) at Stellenbosch University, and the Natural Resources Institute at the University of Greenwich, developed a mathematical model, which showed that recent increases in temperature could account for the simultaneous decline of tsetse. The results provided evidence that locations such as the Zambezi Valley in Zimbabwe may soon be too hot to support tsetse populations.

“If the effect at Mana Pools extends across the whole of the Zambezi Valley, then transmission of trypanosomes is likely to have been greatly reduced in this warm low-lying region”, says Dr Jennifer Lord, lead author and postdoctoral fellow at LSTM.

While this would be good news for the disease situation in Zambezi Valley, rising temperatures may have made some higher, cooler parts of Zimbabwe, more suitable for the flies.

Professor John Hargrove, Senior Research Fellow at SACEMA, says the effect of recent and future climate change on the distribution of tsetse flies and other vectors, particularly mosquitoes, is poorly understood: “We don’t know, for example, whether the resurgence of malaria in the East African highlands in the 1990s was caused by rising temperatures or by increasing levels of drug resistance and decreasing control efforts.

“In general, the ways in which climate change will affect the spread of infectious diseases in sub-Saharan Africa is poorly understood because of sparse empirical evidence,” he adds.

However, work on tsetse and trypanosomiasis carried out at Rekomitjie over the past 59 years has produced long-term datasets for both vector abundance and climate change. The research station is located inside a protected area and has been free of agricultural activities since 1958. In 1984, the area was designated a UNESCO World Heritage Site. As not much has changed other than climate, the data from the site provided the ideal opportunity to develop a temperature-driven model for tsetse population dynamics.

Unlike mammals and birds, insects such as tsetse flies cannot regulate their own body temperatures, and their development and mortality rates are therefore strongly influenced by environmental temperatures. Pupae cannot survive at sustained temperatures below 16 or above 32° C. In addition, tsetse populations can become established in an area only if there are sufficient numbers of host animals and suitable vegetation to support tsetse, Prof. Hargrove explains.

He warns, however, that the Hwange National Park in Zimbabwe and Kruger National Park in South Africa are examples of areas where suitable hosts and habitat for tsetse are abundant. “Tsetse flies did occur in these areas in the 19th century, but they were always marginal because the winters there were rather too cold. With the massive rinderpest outbreak of the middle 1890s, when the vast majority of ungulates died, tsetse disappeared from these areas and have never established themselves again. But if temperatures continue to increase there is a danger that they may re-emerge.”

While tsetse-borne disease holds no danger for wildlife, as they have adapted to each other over millennia, control measures might have to be adopted in case tsetse re-occupy these parks and threaten cattle and humans nearby. According to Prof. Hargrove prophylactic drugs can protect livestock from the tsetse, but no such drugs are available for humans. The only sure way of protecting both livestock and humans is to attack the fly.

The paper:

Climate change and African trypanosomiasis vector populations in Zimbabwe’s Zambezi Valley: A mathematical modelling study

Jennifer S. Lord, John W. Hargrove, Stephen J. Torr, Glyn A. Vale

https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1002675

Abstract

Background

Quantifying the effects of climate change on the entomological and epidemiological components of vector-borne diseases is an essential part of climate change research, but evidence for such effects remains scant, and predictions rely largely on extrapolation of statistical correlations. We aimed to develop a mechanistic model to test whether recent increases in temperature in the Mana Pools National Park of the Zambezi Valley of Zimbabwe could account for the simultaneous decline of tsetse flies, the vectors of human and animal trypanosomiasis.

Methods and findings

The model we developed incorporates the effects of temperature on mortality, larviposition, and emergence rates and is fitted to a 27-year time series of tsetse caught from cattle. These catches declined from an average of c. 50 flies per animal per afternoon in 1990 to c. 0.1 in 2017. Since 1975, mean daily temperatures have risen by c. 0.9°C and temperatures in the hottest month of November by c. 2°C. Although our model provided a good fit to the data, it cannot predict whether or when extinction will occur.

Conclusions

The model suggests that the increase in temperature may explain the observed collapse in tsetse abundance and provides a first step in linking temperature to trypanosomiasis risk. If the effect at Mana Pools extends across the whole of the Zambezi Valley, then transmission of trypanosomes is likely to have been greatly reduced in this warm low-lying region. Conversely, rising temperatures may have made some higher, cooler, parts of Zimbabwe more suitable for tsetse and led to the emergence of new disease foci.

My analysis

A model for fly population mortality is only as good as the temperature data used to run the model. It appears they only used one source of temperature data, the only one available to them, the Rekomitjie Research Station.

Interestingly, this helpful photo was also included in the press release from Eurekalert. It is the weather station used to monitor climate at the Rekomitjie Research Station, Zimbabwe. I provide it below, click for full-size.

At the scale displayed above, you might not notice some important details about the weather station itself, but I did. Here it is, magnified:

Notice anything odd? I sure did. As many of you know, I’ve spent years looking at weather stations around the world, spotting problems that contribute to temperature bias. This one has at least four visible biases that will likely cause it to read warmer than it should, especially in the overnight Tmin temperature.

Here are the issues:

1. Metal roof is nonstandard. It looks like they used the same sheet roof as the buildings in the background.Stevenson Screens are defined to have a wooden roof, painted white. I would expect this metal roof to bias the Tmax upwards in days with direct sun.
2. Closer to surface than normal. While I can’t absolutely confirm this with a measurement, it appears the based of the Stevenson screen is about 1 meter from the surface, based on my experience with inspecting weather stations.  The standard is supposed to be a minimum of 1.5 meters. This will bias both the Tmax and Tmin upwards if my observation is correct.
3. Cattle guard surrounding station. This metal structure will act like a heat sink, biasing the Tmin temperature upwards as it dumps the heat it has absorbed from the sun during the day.

We don’t know when these changes occurred in the record, but it is clear to me that combined, these three observed issues at the Rekomitjie Research Station will likely contribute an upward bias effect on temperatures measured by this station.

4. But wait, there’s more.

The view of the station from Google Earth also tells a story. about upwards temperature bias. It is well-known that when a weather station does not have unobstructed air flow around it, it will contribute to upwardly biased Tmin temperatures at night. It is also known that trees around the weather station prevents some Long Wave IR from the earth warmed during the day by the sun from being sent into the upper atmosphere, being reflected back to the ground by tree leaves. This keeps the air near the ground warmer.

As we see in the Google Earth photo below, the weather station is surrounded by trees and structures, in addition to the cattle guard. Below is an aerial view with 100 meter, 30 meter, and 10 meter distance rings plotted, to be compatible with findings for temperature biases in Leroy 2010 ¹, which is a siting standard accepted by the World Meteorological Organization (WMO):

As you can see, there are quite a few obstructions within the 100 meter circle (the largest red one) and several  within the 30 meter circle the smaller red one. The cattle guard is within the ten meter circle, and from the photo, looks to be less than 3 meters (~10 feet) from the Stevenson Screen. Per the specifications in Leroy 2010, that would make this station a Class 5, with up to 5°C uncertainty in the temperatures it records:

Here is what the  Rekomitjie Research Station temperature plot looks like, from the paper, figure1:

As you can see in Figure1 from the Lord et al. paper² , the span of temperature anomaly from 1965 to present is about 1.5°C, which is still smaller than the uncertainty of a Class 4 station at 2°C, or considering the cattle guard, making it a Class 5 station, an uncertainty of 5°C.

Due to the siting problems, the uncertainty swamps the signal, no matter how you look at it, rendering the claims made from the data to be meaningless.

I don’t particularly blame the authors or the reviewers for not noticing this problem, because they aren’t climatologists or meteorologists they are doctors and entomologists, who wouldn’t even know to look for these sorts of problems.

However, I can blame them for this, from their own press release:

The study is based on prolonged laboratory and field measures of fly densities from the 1990s, and nearly continuous records of climatic data since 1975, recorded by researchers based at the Rekomitjie Research Station in the park. Since the 1990s, catches of tsetse flies from cattle in the park declined from more than 50 flies per animal per catching session in 1990, to less than 1 fly per 10 catching sessions in 2017. Since 1975, mean daily temperatures have risen by nearly 1° C and by around 2° C in the hottest month of November.

All well and good, assuming they actually had good climate data (they don’t), but then there’s this from A Brief History of Tsetse Control Methods in Zimbabwe and Possible Effects of Climate Change on Their Distribution

Bold mine

Odor Baited and Insecticide Treated Targets
It has been shown that the low reproductive rates of tsetse mean that the kill rate needs only to be relatively low in order to have a major control effect (Hargrove, Torr, & Kindness, 2003 ⁴).

….

The application of insecticides directly to cattle was re-instated in the 1980s and 1990s. While the technique had been used since the 1940s, improvements in chemicals and application techniques, as well as improved understanding of fly behavior, have seen this approach yield impressive results (Hargrove et al., 2012; Torr et al., 2011; Torr et al., 2007; Hargrove et al., 2003).

The insecticide can be either be applied as a dip spray or as a pour-on formulation. The pour-on approach, applied monthly, is less error prone, and has been proven more flexible and adaptable in more remote regions, while allowing herders to adapt the approach as necessary (Swallow et al., 1995). However, this pour on method is relatively costly. The lower cost of the dip spray, and the ability to combine it with tick control, makes this a very cost-effective measure to curb AAT (Chadenga, 1992).

Despite these various control measures, neither tsetse nor trypanosomiasis have been eradicated in northern Zimbabwe. Recently, there was a spate of new cases of HAT in northern Zimbabwe, and a number of new programs and initiatives are underway to address this issue (Scoones, 2016). With the radical changes in rural livelihoods and settlement patterns that have occurred in Zimbabwe since the start of the fast-track land reform program in 2000, “it is still unclear how the reconfigured land use and occupation structures have changed exposures to trypanosomiasis” (Dzingirai et al., 2013). In addition, the potential long-term affects of climate change have also been unclear. Changes in climate could dramatically impact the fly belts, either enlarging or reducing them, depending on the changes that take place, and how these affect tsetse population growth rates and habitations.

With such low kill rates by insecticides having a “major control effect”, and the costs of insecticide control steadily decreasing with “impressive results* (Hargrove et al., 2012 ⁵; Torr et al., 2011; Torr et al., 2007; Hargrove et al., 2003 ⁴), one wonders if the reduction is Tsetse flies has any connection to “climate change” at all.

If it were me, I’d be withdrawing this paper as being unsupportable by the uncertain temperature data alone. I think they set out to show “climate change” was a factor, but didn’t bother to really look at the data uncertainty, nor the true impact of control measures.

UPDATE: As pointed out in comments, I missed something important in my initial review. Note this:

It has been free of agricultural settlement since 1958, when the people living there were relocated [22]. Since then, the combined area has been protected against settlement, agriculture, and illegal hunting and logging and was designated a UNESCO World Heritage Site in 1984. In this area, HAT occurs, and tsetse populations have not been exposed to any form of control.

So there’s no insecticide treatment involved, at least not inside the park. There’s also this:

Sampling of tsetse at Rekomitjie, in pursuit of various ecological and behavioural studies, has suggested a decline in tsetse abundance in the last two decades. It is difficult to interpret the catches confidently because they have been made using widely different methods at irregular intervals. From 1966, however, fed female G. pallidipes have been collected from stationary oxen at Rekomitjie, with the sole original aim of providing test insects for bioassays [31–36]. Because these collections were made using a single sampling system, run at approximately the same time each day, the change in the numbers collected offer an indication of the extent of the decline in tsetse abundance over recent decades.

Catches were made for 3 hours in the afternoon during the period of peak tsetse activity [42]. Each collection team comprised two hand net catchers and an ox, operating within 2 km of the research station. Each team operated at least 200 m from other teams, in areas chosen to maximise catches in accord with seasonal changes in the distribution of tsetse between vegetation types [43]. In the 1960s, it was usual for each team to take enough tubes to collect a maximum of about 50 flies each day. This quota was set in consideration of the minimum expected catch at that time and has been maintained at this level ever since, even though it has proved impossible to meet the quota in the last two decades. Daily records are available from 1990 for the number of catching teams employed, and for the catch of each team. The monthly averages of the number of flies caught per team per day are taken as indices of fly abundance. Prior to 1990, tsetse catches regularly reached the upper limit of 50 flies; thereafter, this hardly ever occurred. Fitting the model only to catch data for the period after 1990 ensured that there was no truncation of data used in the fitting procedure.

They cite a reference [42] A diurnal and seasonal study of the feeding activity of Glossina pallidipes Aust.

which says:

In a continuation of studies of diurnal and seasonal feeding activity of Glossina pallidipes Aust. in thick riverine vegetation at Rekomitjie, in the Zambezi Valley, Southern Rhodesia, flies attracted to a stationary black ox were allowed to become engorged, then caught, marked and released.

…

From comparison of catches off a moving and a stationary ox, two stationary oxen standing together, and two standing out of sight of one another, in each case one being red and the other black, it appears that colour of the bait-animal is not of great importance as an attractant to G. pallidipes unless a definite choice is presented, when flies show preference for feeding on a dark surface.

So, they have a stationary “test oxen” they presumably tie up somewhere, and a pair of oxen for a color test. This seems to me to be a ridiculous sampling method, because in the actual farming situation there in Zimbabwe, cattle would be in groups, free roaming around, and foraging. They’d be rubbing into bushes and trees in the process. From the Rory Pilossof paper³, we learn this about the behavior of Tsetse flies:

Ground-spraying worked on the basis that tsetse flies spend much of their life “resting in cool, shady places provided by trees [and] holes … and directing a persistent insecticide at these sites should achieve a good control measure”.

So, with flies spending much of their lives ““resting in cool, shady places provided by trees [and] holes…” and with a “stationary ox” and a pair of oxen providing the only sustenance for blood sucking flies, you can see that it’s no wonder their catch rate has diminished so dramatically. The cattle is mostly gone from the test area, and the test animals with “catch teams” don’t likely even get to roam around where it can brush up against or take sun and heat shelter near trees and holes the flies frequent. Then one wonders if the “catch team” isn’t doing the work while wearing insecticide themselves to keep from being bitten. Did they wear the same color clothing for years to prevent the flies from being given a choice of a light and dark surface? It would seem the catch teams provide a potential bias themselves. And where are the control oxen that aren’t with catch teams?

The authors of this paper haven’t gathered any of their own data, but instead rely entire on single station weather data and fly catch data gathered by others, with no obvious control group for fly catch or weather data..

This paper is even more ridiculous than I had originally thought.

In my opinion, the lack of cattle hosts in the park and the fly capture method itself is the reason for the decline in flies, not climate change. Flies don’t thrive without cattle, and cattle don’t get bitten as much when they are led around by “catch teams” who would likely by human nature not want to get bitten. Even if the temperature data wasn’t suspect, correlation isn’t causation here.

References: 

1. Leroy, M., 2010: Siting Classification for Surface Observing Stations on Land, Climate, and Upper-air Observations JMA/WMO Workshop on Quality Management in Surface, Tokyo, Japan 27-30 July 2010 http://www.jma.go.jp/jma/en/Activities/qmws_2010/CountryReport/CS202_Leroy.pdf
2. Jennifer S. Lord et al., Climate change and African trypanosomiasis vector populations in Zimbabwe’s Zambezi Valley: A mathematical modelling study https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1002675
3. Rory Pilossof A Brief History of Tsetse Control Methods in Zimbabwe and Possible Effects of Climate Change on Their Distribution International Journal of African Development v.4 n.1 Fall 2016 https://scholarworks.wmich.edu/cgi/viewcontent.cgi?article=1089&context=ijad
4. Hargrove, Torr, & Kindness, 2003  Insecticide-treated cattle against tsetse
   (Diptera: Glossinidae): What governs success? Bulletin of Entomological Research,
   93(3), 203-217.
5. Hargrove, J. W., Ouifki, R., Kajunguri, D., Vale, G. A. & Torr, S. J. (2012). Modeling the
   control of trypanosomiasis using trypanocides or insecticide-treated livestock. PLoS
   Neglected Tropical Diseases, 6(5).