The Unpaid Garbage Man : An economist takes a look at aquatic litter

I am an unpaid garbageman, one among thousands world wide. Tired of swimming in the garbage from city streets we struggle to make sense of and stop the continuous pollution of our water ways. We exchange data, experiences and ultimately we connect our local projects to a global problem. A sort of informal citizen science project that grows in sophistication using open-data, scientific publications, DIY biology and personal observations to demonstrate the absurd. These are the preliminary results of one project that was born out of frustration, revulsion and ultimately love; love for our environment, the ocean and our children.

More than 26' 000 items (trash) were removed, counted and identified by three people on a small portion of the shoreline of Lake Geneva, Switzerland. The volunteer trash removal operations followed a written protocol between November 2015 and August 2016.

By following a protocol such as the Marine Litter Watch, a data rich environment is created that can be used to relate “trash on the shoreline” to publicly available economic, demographic and environmental data. Coupled with advances in open source technology and the increased availability of “Automated Program Interfaces” (APIs), in our case “Open-Data-Kit” (ODK) and Google Charts, small unfunded operations streamline data gathering and publicly display results in an easy (some coding skill required) and cost effective manner (almost free).

Chart 1 : Data from the first project Nov 2014 — Aug 2015. Collected garbage was measured in liters, samples were taken from the river mouth at the “Baye de Montreux”.
Chart 2 : Data from the second project Nov 2015 — Aug 2016. Collected garbage was measured in pieces, samples were taken from the river mouth at the “Baye de Montreux”.

Chart 1 for example displays the daily total of trash removed and the monthly totals of hotel arrivals in Montreux for November 2014 to August 2015. the litter data comes from our surveys and the hotel check in data comes from the federal office of statistics.

Hotel Check Ins represent change in population density over a short period of time at a specific location. In Switzerland there is a “Taxe de séjour” that is added on to the cost of each hotel room. Furthermore each “occupant” will at a minimum eat and use the bathroom, thus increasing revenue from value added tax and putting pressure on public utilities.

That the pattern is repeated the next year is perhaps just a coincidence (chart 2).

I am an unpaid garbage man with modern data tools and I understand that the trash on the beach is not a random occurrence.

Assuming that each piece of trash found on the shoreline is independent of any other piece of trash on the shoreline and the the occurrence of any item on the beach is random then the multiplication rule for probability can be used. Each piece of trash belongs to one of 132 usage categories and one of five material categories, so 5*132 = 660 different combinations for each entry. Therefore the chances of finding any item on the list is 0.0015 (1/660), take it a step forward and still assuming a random-discrete distribution the chance of finding the same item twice is 2.27*e-6 (1/660*1/660). But our data shows that some objects make up almost 30% of the items found .

For more information about random discrete probabiliy : see Dartmouth teaching aids or Wikipedia has a pretty good explanation. IBM Big Data University is also free and good !
I am an unpaid garbageman when I plot the quantity of trash removed against independent variables a pattern emerges year over year.

Obviously there are other factors at play and the garbage found in the water and on the shoreline is not random (so we can’t use the multiplication rule!). When the data is put in a “time series” peaks and troughs become evident. Peaks and troughs can be caused by a variety of factors such as temperature, weather or increased human (economic) activity near the shoreline. Chart 3 (garbage was counted in liters) and 4 (garbage was counted in pieces) display what appears to be a repeating pattern year over year, indifferent of the way the garbage is quantified.

Chart 3 : The first year the litter was measured in liters. In the second year less samples were taken but each piece of trash was counted and identified according to MLW criteria.
Chart 4 : Daily totals of trash collected measured in pieces.

The increases in economic activity represent profits, increases in value added tax revenue, hotel tax revenue and a temporary rise in payrolls. The resulting rise in profits also generates annual tax revenue when accounting profits are taken into consideration.

The fluctuations of trash collected per day are subject to many independent variables. In charts 3 and 4 the peaks of garbage are consistent with temperature change, in charts 1 and 2 they are consistent with hotel nights.

That this pattern is repeated year over year is perhaps just a coincidence or it suggests that more sophisticated data analysis techniques are required to account for other, unidentified independent variables that may be at work.

I am an unpaid garbageman that regularly removes hazardous material from the environment. That material is accompanied by a variety of biological and chemical contaminants from human activity. The increase in material found is the result of increased economic activity and its presence is a clear violation of the law.

According to article 4 of the Swiss “Act on the protection of water” pollution is defined as “Any harmful alteration of the physical, chemical or biological properties of water”. There are many studies that relate plastics (80–90% of what we find) to bio-accumulation of persistent organic pollutants (POPs) and presence of endocrine inhibitors(Teuten et all, 2009).

Image 1 from NOAA marine debris program

Endocrine inhibitors and POPs have been linked to numerous adverse health effects in all organisms. These effects range from slow growth rates to adverse sexual maturation of aquatic organisms. Consequently, these effects have long term and wide ranging significance to the ecosystem and the survival of species dependent on a healthy ecosystem(Thompson et all, 2009). Furthermore, ingestion of solid waste (primarily plastics) has been directly linked to adverse health effects in all organisms that inadvertently consume plastics, indifferent of the chemical composition of that plastic and the organism(image 1).

In the latest report on “The state of Swiss water ways” (french version) the “Federal Office of the Environment” identifies medications and sewage effluents as important contributors to the degradation of water quality in small to medium size streams. In an interview on live radio with the authors of this article the director of “Surface water quality” for the Canton of Vaud evokes a water treatment system that is undersized and in constant revision. (link to interview in french)

Visible clues, such as the presences of cotton swabs or tampon applicators, can be a good indicator of prior contamination by untreated sewage. Presence of these items was noted at 12% of the survey sites in the latest report of “The state of Swiss water ways”. Preliminary results for this survey indicate that depending on the location cotton-swabs, tampon applicators and feminine hygiene pads contribute between 4% and 16% of the trash removed from the beach.

Image 2 : 4mL of lake water after 24 hours of incubation. Montreux July 12, 2016.
Chart 5 : Totals of trash and number of E.coli colony forming units June 2016 — July 2016.

However in the “System of analysis and apprection of Swiss Lakes” there is no mention of any metric to evaluate the density of those items in the environment.

The best low tech way to identify the presence of raw sewage in the water is to sample the water and culture those samples in a chromogenic medium to identify the micro-organisms present, specifically E.coli.(image 2)

Samples were taken and cultured from three different sites every Tuesday for six weeks from June 06, 2016 to August 09, 2016.

Chart five displays the results of the cultures and the daily totals of trash collected for June and July 2016.

Like trash, there are many independent variables that can explain the spike in E.coli for the period, increased human activity is just one of those.

I am an unpaid garbageman. The cost to remove the hazardous objects from the environment follow a predictable model. Although there is no demand for my services I have calculated a price.

Using the meta data gathered during the surveys it is possible to calculate time on site, admin time and incorporate other costs such as server time, software licenses, transportation, equipment and communication. Once a salary and a “fair” benefits package is included this cost can be expressed in price/pieces of trash.

Chart 5 : Price per piece versus daily total. Greater supply (blue line) = lower price per unit. At the intersection 0.36CHF/297pieces the density is 4.71 pieces per meter of shoreline, or half the weighted average for the entire project. Is this the definition of a clean beach?

The cost for removal of one piece of trash ranges from CHF 0.22 — CHF 0.87 per unit. The cost is largely dependent on the density of trash on site. In this way it fits a very simple and familiar model : greater supply = lower price (chart 5). There are a few other variables associated with the cost such as experience, weather conditions, interactions with the public and the type of terrain.

I am an unpaid garbageman. By classifying and counting the objects I remove from the beach and using simple data analysis techniques I can identify the source of many different types of diffuse aquatic litter.

Zero is how many times a survey operation had a total of zero. Of the 132 official Marine Litter Watch categories 89 are represented in the current survey for Montreux. Of the 43 unused categories 20 are from the commercial fishing industry and 6 represent heavy industry neither of which can be considered “very important” economically in the region.

Typical amount and type of trash found on the shoreline. Plage de l’ Arabie August 25, 2016, beach length = 18meters. The complete inventory is available in our toolbox

Some objects are found sporadically , other objects are found at almost every clean-up operation.

Combined the most common items create a reference point or baseline that is never zero. Variations from the baseline represent specific insults to the environment. The nature and the source of the insult can be approximated by identifying the individual items that deviate from the baseline.

Preliminary results suggest that the density of trash on Lake Geneva is twice that of the shoreline in the United Kingdom (results for 2015 “Great British Beach Clean”) and almost equal to that of France (link to OSPAR data). The current weighted average for our study is 8.25 pieces of trash per meter. There are about 142 kilometers of coast on the Swiss side of Lake Geneva, 142*1000*8.25 = 1’171’500 pieces of trash can be found on any given day on our coast line.

According to Marine Litter Watch, NOAA, United Nations and OSPAR the trash found on the shoreline is a small representation of what is in the water. In other words it is safe to say that Switzerland is exporting around 1'171'500 pieces of trash to France on any given day. In exchange for all this trash the French get a robust border economy with one of the world’s richest countries.

I am an unpaid garbageman in one of the worlds richest countries; the least paid element of a market economy that has not yet understood the value of the environment. The trash that I don’t remove will go downstream simultaneously collecting and releasing molecules that will harm all organisms. At its final destination the trash will choke, mutilate, poison or entrap a multitude of species thousands of kilometers away. All levels of government are aware of the problem and refuse to engage in any meaningful effort to measure the effects or density of these pollutants in our environment. The work I do has value and I have put a price on that value.

For more information :

Notes :

(1)This is the second project, the first project (Nov 2014 — Aug 2015) was counted in liters and consisted for the most part of samples from two beaches. For this project we have 17 different locations giving a better representation of the litter found along the shoreline.

(2)This is an independent and unsupported project, the authors received no financial support from any institution (public or private) for the litter removal operations over the past two years.

(3)The supplies and technical oversight for the microbiological surveys was offered by three organisations : AGIR! Hackuarium and

(4)We are approaching the end of the sampling period, we will mash our data with the available public data to demonstrate the obvious. The finished report will be done October 15, 2016.

(5)We have requested monthly totals for “Value added tax” and “Expenses related to maintaining the Environment” from the city of Montreux and the federal government.

(6)Daily totals for precipitation, max wind values and the average daily flow rate for the Rhone river have been requested.

Special thanks