Agent-Based-Modelling. Pond eutrophication in agroecosystems and the influence of combinations of pesticides and fertilizers on aquatic productivity


Term Paper, 2018

23 Pages, Grade: 1,3


Excerpt


List of Contents

Abstract

1. Introduction

2. Material and Methods
2.1 Purpose
2.2 Entities, state variables and scales
2.3 Process overview and scheduling
2.4 Design concepts
2.5 Initialization
2.6 Input data

3. Results

4. Discussion

5. List of References

Attachment

Abstract

The paper asks, if ponds located in agroecosystems can be prevented from tipping over by variation in combinations of fertilizers and pesticides. Through a simulation in NetLogo 6.0.2., an agent-based programmable modeling environment, the interplay of two pesticides (Cypermethrin and Malathion) and a fertilizer (Nitrogen + Phosphate) and their effect on the aquatic productivity is observed. The model depicts an agroecosystem with three separate fields and two ponds. The model is simplified by assuming consistent environmental conditions and narrowing down the pond population to one zooplankton species and one phytoplankton species (Chlorella vulgaris). Through variation of variables different scenarios are generated.

The research revealed that input of pesticides – even without fertilizers which are known for supporting phytoplankton growth – results in eutrophication through decreasing zooplankton population and therefore blooming phytoplankton population.

The original contribution of the simulation is in revealing that beyond the amount of chemical substances used in farming, the variation of input is also to consider, since certain ones may have greater negative effects than assumed so far.

1. Introduction

The use of synthetic fertilizers and pesticides is increasing in modern agriculture, contaminating whole agroecosystems. Pesticides that are applied to protect crop from weed and pest insects also expose non-target organisms to higher risk. Since most of the pesticides are persistent in the environment, surrounding flora depauperates and several faunal species are deprived of their food resources. (Umweltbundesamt, 2017) (Schachtschabel et. al, 2010, S.114). Only a small amount of applied pesticides has the desired effect and hits the target. The rest gets into the soil and air and, through run-off, into nearby water bodies and negatively affect their health, structure and function. Thus, it can be said that pesticides are, together with excessively applied fertilizers, the main pollutants of aquatic ecosystems (Hedge et. al, 2014, S. 435).

Pesticide decline the zooplankton population, which results in the bloom of phytoplankton populations, such as algae. Fertilizers in turn, enhance the phytoplankton productivity. Thus, both agrochemicals lead to increasing population of phytoplankton, which is a main cause of eutrophication (Hedge et. al, 2014, S. 435) This phenomenon forms the basis for this research.

While there is a lot of research on the single effects of pesticides and fertilizers on aquatic ecosystems, there are only a few studies on the effects of their combination. Therefore, the aim of this paper is to examine the effects of combinations of pesticides and fertilizers used in agroecosystems on nearby water bodies. Through a model, created with NetLogo 6.0.2. and use of several data, an agroecosystem consisting of three fields and two ponds is simulated.

The model is based on a study by Hedge et. al, who tested the influence of widely used pesticides and fertilizers on chlorophyll and zooplankton production for rice paddys in Western Ghats Regions of India. The experiment was conducted for 42 days by using a laboratory mesocosm. As stated in the study report, environmentally realistic concentrations of the pesticides Malathion and Cypermethrin and the fertilizers Nitrate-nitrogen and Phosphate-phosphorus were used. The substances were tested in different combinations such as “pesticide-pesticide” and “pesticide-fertilizer” (Hedge et. al, 2014, S. 434).

Thus, in the model the two pesticides are assumed to be Cypermethrin and Malathion, of which one is less toxic than the other. The fertilizer is assumed to be a mixture of Nitrogen and Phosphate at same concentrations.

At first, the modelling process will be outlined by explaining the profound concepts, the used data and the initialization in detail. Afterwards the results of two different simulation scenarios will be presented. Then follows a discussion of the model, the process and the found results.

2. Modelling Process: Material and Methods

2.1 Purpose

The purpose of the model is to explore the effects of combinations of pesticides and fertilizers used in agroecosystems on the aquatic productivity and with that, the change in the trophic level of ponds located nearby. The model is not to explain the general process of eutrophication but should observe the interplay of certain pesticides with fertilizers and how the populations of zooplankton and phytoplankton are affected by that. The model furthermore tries to examine how, through variation in concentration of pesticides and fertilizers, conservation measures of ponds could look like and how different they work.

2.2 Entities, state variables and scales

The computation of this simulation comprises four components; pond initialization, agent initialization and farmer initialization. The basic time step represents one day.

Entities within the model include agents representing the zooplankton, phytoplankton, initial nutrients, fertilizers, pesticides, farmers and the landscape. The spatial unit includes the following landscape data (areas): soil, fields, ponds and sky.

Agents/individuals:

Zooplankton: state variables: energy, affected?

Sub-type: Small zooplankton with energy < 15

Phytoplankton: state variables: energy, age

Initial nutrients (natural nutrient input): no state variables

Fertilizer: state variable: presence in water

Pesticide A (toxic): state variable: presence,

Pesticide B (less toxic): state variable: presence

Spatial units:

Pond areas: state variables: nutrient content, O2-concentration. There are two separate ponds which represent the habitat for zooplankton and phytoplankton.

Field areas: There are three different field areas which mark the farmers’ locations and the source of chemical input.

Soil: The soil serves as a connection between the fields and ponds and constitutes the path of chemical substances into the water.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1 . Model design: Agroecosystem with the spatial units ponds, fields and soil (Source: Own illustration).

2.3 Process overview and scheduling

The following process overview assumes, that all sliders are switched on and all variables are greater than zero.

By starting the simulation, the zooplankton and phytoplankton randomly move around the ponds at the same speed. By moving forward, they lose energy.

If a zooplankton-individual encounters a phytoplankton-individual, it feeds on it. By eating, it gains energy. A small zooplankton grows to a bigger size, as soon as its energy is greater than 15 which in turn is nourished by nutrients (fertilizer). The interdependence of zooplankton and phytoplankton follows the principle of a predator-prey-relationship. If a zooplankton-individual was hit by a pesticide, it state is “affected”, and it dies after four days.

If a phytoplankton-individual encounters a nutrient or a fertilizer, it “eats” it and gains energy. With an age of 80 days, a phytoplankton dies.

When both, zooplankton and phytoplankton run out of energy, they die. To allow their populations to subsist, each of them has a variable probability of reproducing.

By starting the simulation, the farmers apply fertilizer, pesticide A and pesticide B into their fields. These diffuse into the soil, where they randomly move to right and left. If they encounter the pond areas, they do not move out again. Their speed remains the same all the time.

Once a fertilizer-individual is in the pond, it affects the water by increasing its nutrient content by a value of 0.1 in every time step.

Once the pesticides are in the pond, they can affect the zooplankton. If a pesticide-A-individual (toxic) encounters a zooplankton-individual it sets its energy down by 10 and turns its state to “affected”. If a pesticide B-individual (less toxic) encounters a zooplankton-individual it sets its energy down by 5 and turns its state to “affected”.

The simulation stops, if either one of the ponds tips over.

2.4 Design concepts

Basic Principles: Two main principles influence the model’s design: Eutrophication of lentic water bodies through nutrient-input and a predator-prey-relationship in a trophic web system.

Emergence: Population dynamics and pond eutrophication emerge from the behaviour of the individual agents. Excessive fertilizer input and high use of pesticides result in eutrophication quite fast and can be rather seen as a “built in” result. However, the time/moment of pond contamination and instability of population for each pond emerges from the combination of chemical input of the farmers. Since they use different pesticides and are located differently, the direct effect on the ponds is not predictable or certain.

Collectives: Fertilizer, pesticide A and pesticide B are three defined collectives used in the model. The respective individuals all have the same properties which only change collectively.

Interaction: The populations dynamics and individual state variables of spatial units and agents change through interaction between the chemical substances and the plankton. The phytoplankton is directly affected by fertilizers and nutrients, since they increase its energy. It is furthermore affected by the zooplankton, which kills it by eating. The zooplankton is directly affected by pesticides which decrease their energy and cause them to death after four days (simulating bioaccumulation of pesticides) of encounter (cf. Figure 2) (Krautter et. al, 2008).

Abbildung in dieser Leseprobe nicht enthalten

Figure 2 . Interaction: die-from-pesticides procedure (Source: Own illustration).

Stochasticity: Stochasticity is partly used for the reproduction of plankton. A random value of an adjustable range determines the reproduction chance (cf. Figure 3). Moreover, some initial state variables are stochastic (cf. Table 1).

Abbildung in dieser Leseprobe nicht enthalten

Figure 3 . Stochasticity in reproduction procedure (Source: Own illustration).

Observation: To observe the effects on population dynamics and trophic states of the ponds, following data is collected: Growth rates of zoo- and phytoplankton population, oxygen concentration, nutrient content, period and amount of pesticide and fertilizer input.

2.5 Initialization

Initially, the default settings of each of the ponds are defined: the nutrient content, the oxygen concentration, the number of zooplankton and the number of phytoplankton.

Nutrient content: The initial nutrient content determines the nutrition of the ponds at the beginning of the simulation. The maximum lies at 50 mg/l. Measurements showed that 21 % of surface waters in the EU have a nutrient concentration below 2 mg/l and are therefore considered to be extremely healthy water bodies. In only three percent of surface waters, the concentrations exceeded 50 mg/l (Europäische Union, 2010, S. 2). The default nutrient content of both ponds is 2 mg/l.

Oxygen concentration: The oxygen concentration (in %) signifies the initial trophic state of the ponds:

Trophic state 1 (oligotrophic): > 80 %.
Trophic state 2 (mesotrophic): 40 - 80 %
Trophic state 3 (eutrophic): < 40 %
Trophic state 4 (hypertrophic; tipped over): 0 %

The default setting for both ponds is 100 %, which means they are healthy ponds in trophic state 1 (Schwoerbel et. al, 2013, S. 250).

Number of zooplankton: The initial size of zooplankton population in each of the ponds. The default-setting in both ponds is 50 and is chosen arbitrarily. The number can be variated and is not based on data. The number can be variated and is not based on data.

Number of phytoplankton: The initial size of phytoplankton population in each of the ponds. The default-setting in both ponds is 20 and is chosen arbitrarily. The number can be variated and is not based on data.

Furthermore, the reproduction rates of zooplankton and phytoplankton, the zooplankton’s energy gain from food and the phytoplankton’s energy gain from nutrients are defined.

Zooplankton gain from food: The amount of energy which zooplankton gain from eating phytoplankton. The default-setting is 20.

...

Excerpt out of 23 pages

Details

Title
Agent-Based-Modelling. Pond eutrophication in agroecosystems and the influence of combinations of pesticides and fertilizers on aquatic productivity
College
Leuphana Universität Lüneburg
Grade
1,3
Author
Year
2018
Pages
23
Catalog Number
V468575
ISBN (eBook)
9783668949010
ISBN (Book)
9783668949027
Language
English
Keywords
agend-based-modelling, pond
Quote paper
Melanie Bayo (Author), 2018, Agent-Based-Modelling. Pond eutrophication in agroecosystems and the influence of combinations of pesticides and fertilizers on aquatic productivity, Munich, GRIN Verlag, https://www.grin.com/document/468575

Comments

  • No comments yet.
Look inside the ebook
Title: Agent-Based-Modelling. Pond eutrophication in agroecosystems and the influence of combinations of pesticides and fertilizers on aquatic productivity



Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free