• No results found

Adverse effects of agricultural land use on benthic invertebrates in lowland streams

N/A
N/A
Protected

Academic year: 2021

Share "Adverse effects of agricultural land use on benthic invertebrates in lowland streams"

Copied!
23
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

1

Adverse effects of agricultural land use

on benthic invertebrates in lowland

streams.

Bsc Thesis by: Toon Driessen

In cooperation with T. Theirlynck and J. Pasqualini.

Supervisors: P.C. Dos Reis Oliveira & M.H.S. Kraak

(2)

2 Abstract

The Dutch landscape is typically patchy and used for intensive agriculture. Each type of land use is releasing land use specific toxicants into the environment, which are continuously entering lowland streams. These toxicants accumulate in sedimentation areas, affecting the benthic organisms living there. Due to the multi-stress nature of the environment it is difficult to relate observed effects to specific stressors. Therefore this research aimed to assess the effects different types of agricultural land use on benthic communities. To this purpose we performed a sediment quality triad, consisting of a chemical analysis, a toxicity test and a The area of study was the 'Hierdense Beek' catchment area. Results showed significant differences in nutrient composition. The intensive grassland had highest nitrogen values whereas the corn field had the highest phosphorous levels. The community composition between the forest and cultivated land was also very different, showing high abundances of sensitive species in the forest in contrast to the corn and intensive grassland. The toxicity test did not result in significant differences. Combining the multiple lines of evidence, it is likely that the differences found in community structure are related to the different types of agricultural activity..

KEYWORDS: Sediment quality, Triad, agricultural land use, benthic invertebrates, H. azteca

Introduction

The Dutch landscape is typically patchy and cultivated, more than half of the country is being used for agricultural activities (CBS, 2015). This has a major influence on the remaining natural systems, such as low land streams which flow through these highly artificial landscapes, resulting in less than 4%

of streams with a natural hydro-ecology left in the Netherlands (Verdonschot & Nijboer, 2002).

Rivers and streams receive continuous inputs from agricultural activities, mainly after rain events, which contain a mixture of pollutants. These mixtures are land use specific and therefore differ from site to site (Strayer et al., 2003), resulting in land use specific effects on biota. In addition, streams are channelized to serve agricultural activities better. This is resulting in higher stream velocity, higher peak flows and increased peak flow frequency. Changes in riverbank characteristics by removing vegetation are very common, resulting in less riverbank- and in stream heterogeneity (Allan, 2011). One last stressor is the increase in sediment runoff (Strand & Merrit, 1999), which is thought to be one of the major stressors for benthic invertebrates (Wagenhoff et al., 2012).

Even though water quality standards must be met and many pollutants occur in low concentrations, these pollutants are still thought to pose a risk to the environment. This can be caused by the persistence of pollutants or by continues input of non-persistent pollutants. Many of these pollutants become attached to fine sediment or organic particles, which are accumulating in the sediment in low stream areas (Halling-Sørensen et al., 1998; Muñoz et al., 2009; Wolfram et al., 2012a). This causes benthic organisms living here, to be chronically exposed to a mixture of pollutants. (Daughton & Ternes, 1999; Wolfram et al., 2012a). Pollutants that are not persistent gain fewer attention and are often not on the priority list, although these might do as much damage as persistent pollutants due to the continuous input into the environment (Daughton & Ternes, 1999; Fent et al., 2006).

(3)

3 These combined stressors are creating a multi

stress environment for benthic organisms living in these streams. Since stressors have the potential to interact and the effects are non-linear, the result is difficult to predict (Wagenhoff et al., 2012). Since the sediment is the main environmental compartment of concern, a good method for indicating single or multi stress in natural systems is the Sediment Quality Triad (SQT). This method was developed by Chapman (1990) in order to provide ecologically relevant information on sediment quality using multiple lines of evidence. The method consists of a chemical component, in which the presence of potential toxicants is measured, a bioassay, in which the bioavailability and the toxicity of the sediment sample is measured under laboratory conditions, and a community assessment, which gives meaningful results on the effects in the field. Apart these assessments may give some information, but are not comprehensive. Yet, together these lines of evidence provide a robust and comprehensive assessment of the effects of single as wel as mixtures of toxicants under natural conditions (Chapman, 1996; Chapman, 1990; Leslie et al., 1999; Wolfram et al., 2012b)

The aim of this study was to assess the effects of agricultural land use on benthic biota in lowland streams, using the Sediment Quality Triad. Three types of land use were assessed, a forest, a landscape determined by intensively used grassland and a landscape determined by corn cultivation. It was hypothesized that there will be a land use specific effect on the survival and assembly of benthic communities, and that each type of agricultural land use releases a specific mixture of toxicants into surrounding streams. The forest sample was be used as a reference sample, since this is of major importance in distinguishing effects from local sediment

conditions from effects from anthropogenic pollutants (Chapman et al., 1997).

Allan (2011) and Strayer et al., (2003) already revealed that there are differences in runoff and toxicity between pasture landscapes and crop field. This is in line with the expectation that differences in toxicity would be found. In addition De la Torre et al. (2000) showed the direct toxicity of pig manure to Daphnia magna. If land use has a negative effect on stream biota, it is expected that despite large efforts, restoring natural systems in the Netherlands will be difficult due to the fact that most land is intensively cultivated. It is however important to provide evidence for local and land use specific conditions.

Materials & Methods Sampling locations

Sediment and water were collected from three locations in the 'Hierdense Beek' catchment area (Figure 1). Site one (S1) is a forest stream, this is part of the catchment that is relatively unaffected by agriculture nor surrounded by agricultural activities. Site two (S2) is a stream that flows through a landscape that is mainly used for cultivating corn and site three (S3) is a stream that flows through a landscape that is intensively used for cultivating cattle and consists of grasslands. Sediment and water samples were taken on the 28th and 29th of April. The community samples were taken on the 9th and 10th of May.

Sampling methods

A total number of five sediment and water samples were taken from each location over a stretch of 15 meters, in 5 different siltation areas. The in situ physical parameters current velocity, stream dept, width, substrate,

(4)

4 habitat types were also recorded. Only the

upper two centimetre of the sediment were sampled with a core, since the test organism inhabits this part of the sediment (Borgmann et al., 2005). Sediment and water samples were treated and collected according to the EPA protocol (Plumb, 1981). Samples for analysing community composition were collected by two methods, a surber with a mesh of 0.25 mm and a standard handnet with a 25 X 25 openings (European Standard EN 27 828). The surber was used to sample five sedimentation areas along the 15 meter stretch and the handnet was used to sample all micro habitats present.

Chemical analysis Water analysis

All water samples were filtered through a 0.2 micron filter before analysis. Total nitrogen (Ntot), dissolved organic nitrogen (DON), nitrate & nitrite (NO4 + NO3), ammonium (NH4), orthophosphate (PO4) sulphate (SO4) and chlorine (Cl) were measured with the Skalar SAM ++ auto analyser using the software Flow Access v3.1. Total phosporous (P), sodium (Na), magnesium (Mg), calcium (Ca), manganese (Mn), silicon (Si), iron (Fe), aluminium (Al) and potassium (K) were measured using an Perklin Elmer ICP-OES 8000. Furthermore pH, dissolved oxygen and conductivity were also measured in situ, using a Hach HQ440d multi meter. Flow velo and phisical description

Sediment analysis

Sediment samples were dried and prepared for nutrient analysis by sieving trough a 2 mm sieve before finely grinding for 5 minutes on 400 rpm using the Fritsch Pulverisette 5. The sediment was analysed for macro nutrients (N, C, P and S). Nitrogen, sulfur and carbon were measured using an Elementar Vario EL analyser. Organic matter content was determined by loss of ignition over 16 hours at 375 C°.

A linear model was applied to assess the effects of sediment- and water chemicals on survival, weight and length. Differences in chemical concentrations between treatments were tested by performing an ANOVA for parametric parameters and a kruskal-wallis for non-parametric parameters, both with a post-hoc Tukey-Kramer test. Analysis were done using the software R-Studio (R Core Team, 2014), version 0.98.501.

Figure 1, Detaild map of the Hierdense Beek catchment, indicated with a yellow dot on the full map of the Netherlands. S1 is the Forest site, S2 is the Corn site and S3 is the Intensive Grassland site

(5)

5 Bioassays

The test organism of the 28-day bioassay was the amphipod Hyalella azteca, originating from a culture in Wageningen. H. azteca is an easy to culture, standardized test species. It is sensitive to pollutants, but it is able to coop with varying laboratory conditions (Artlett et al., 2013; Borgmann et al., 1989; Ingersoll et al., 2015; Ingersoll et al., 1995). It lives in and on the sediment, it forms a major component in the food web and represents natural benthic biota (Ingersoll et al., 1995).

The test was performed according to EPA guidelines (Cincinnati, 2000), with slight modifications. General test conditions are shown in Table 1. 24 hours before the start of the test 30 grams of sediment was homogenised in 150 ml beakers and 120 ml overlying water, originating from the sampling locations, was added, resulting in a 1:4 ratio sediment:water. Aeration was applied and food was added. 5 % of sediment dry weight finely ground Urtica dioica was added as food source. There were five replicates per location and one lab control with artificial sediment (OECD, 2004) and Dutch Standard Water was included in the experiment. 10 organisms were added to each beaker at the start of the test. Conductivity, pH and dissolved oxygen were measured every week and hardness before the start and in the end of the test. At the end of the test the sediment and overlying water was filtered through a 150 and 350 µm sieve in order to collect the living and dead animals. Immobile or missing organisms were considered dead. The end points of the test were survival, length and dry weight. Length was determined by making a microscope picture after which length could be measured using imageJ 1.51d. Dry weight was measured by freeze drying the animals first, after which an analytical balance was used to measure the weight in grams. Animals

were classified as offspring if they were smaller in length than the initial average at

start. Differences between treatments were assessed by performing an ANOVA.

Community Composition

Macro invertebrates were sampled at the three sites, using the surber and hand net. The living animals were sorted within 48 hours after sampling and saved in alcohol, after which they were identified to the lowest taxonomic level possible. The following indices were calculated:

Shannon Diversity Index

Table 1, General test parameters and conditions according to USEPA, with slight modifications

Parameter Conditions

Test duration 28 days Temperature 20 ±1°C Light quality Wide-spectrum

fluorescent lights Illuminance ~500-1,00 lux Photoperiod 16:8 h light:dark Sediment:overlying water volume 1:4 Test chamber 150 ml Age of organism 1- 9-day old Organisms/chamber 10

Number replicates 5

Feeding 5% of sediment dry weight ground Urtica dioica

Aeration > 60% oxygen

saturation

Water quality Hardness, at start and end, pH, conductivity and dissolved

oxygen once a week End point Survival, dry weight.

length

Test acceptability Minimum control survival 80%

(6)

6 Where D is the Shannon diversity, and p the

proportion of animals belonging to the ith species.

Margalef Index

S indicates the species richness, and N the total number of individuals counted.

Jaccard Simmilarity Index

The Jaccard Similarity Index was used to compare the community composition of different sites with each other. A is the total number of species in common between b and c, b is the total number of species unique to group b and c is the total number is species unique to group c. It represents the percentage of species that occur in both samples.

A is the number of species in common, b is the number of species unique to sample 1 and c is the number of species unique to sample 2 EPT Index

N is the total number of individuals counted, E is the total number of Ephemera, P the total number of Plecoptera and T the total number of Trichoptera.

Oligochaeta- Chironomid index

N is total number of individuals counted, O the number of oligochaetes, C the number of Chironomids.

Belgian Biotic Index (BBI)

The BBI is calculated using table 4.2 and 4.3 in 'Macro-invertebraten en waterkwaliteit' (Pauw, de & Vannevel, 1993). Each taxon is given a rating, which is high for sensitive species and low for pollution tolerant species. Results

Chemical analysis Water analysis

Water analysis resulted in several significant differences between treatments. The full list of results is presented in Appendix 1. Table 2 contains the location means with the standard error. Between locations, significant differences (p< 0.05) were found for S, Fe, Mn, Mg, Ca, Al, Na, K, NO3, NH4, N, PO4, Cl and SO4, these results are presented in Appendix 2. Because water quality satandards from the Water Framework Directive, extracted from Evers et al., (2012). were available, the found measured average concentrations were compared to these levels (Table 3).

Table 3, Environmental quality standards for good water quality according to the 'Water Framework Directive', from Evers et al (2012).

Compound Forest Intensive Grass Corn N <2.3 5.86 10.40 3.48 P <0.11 0.034 0.029 0.059 Cl < 40 11.83 16.99 8.50 NH4 < 0.15 0.014 0 0.079 NO3 < 0.15 17.29 39.58 6.01 SO4 < 0.15 20.36 24.84 17.44 Nitrogen exceeded the values for good water quality, which has a maximum of 2.3 mg N/l (p < 0.01, V=120), see Table 4 and Figure 2, as well as NO3 (p< 0.01, V=120). NO3-NO2 measurements were treated as if all were NO3, since NO3 is considered to be representative for NO3 + NO2 concentrations (In, 2013). Conductivity differed significantly between sites (p< 0.01, F= 63.39), resulting in the lowest conductivity for forest (mean= 278.2 µS/cm, SE=9.56), then intensive grass (mean= 347.2, SE=4.33) and the highest value for corn (mean=372.6, SE= 1.69), all significantly different from eachother which is shown in Figure 3. pH dit not differ sifnificantly between sites (p=0.16, F=2.13), neither did flow velocity (p=0.61, F=0.50).

(7)

7 Percentage DO did differ between sites,

resulting in a significantly higer value in

intensive grass than compared to corn (p<0.01) and forest (p<0.01) (see Figure 4).

Figure 2, Concentration of Ntot found in the water samples (mg/l) with standard error, compared to the environmental quality standards from STOWA, based on the Water Framework Directive (Evers et al., 2012)

Figure 3, Conductivity measured in situ. Significantly different between all locations. Error bars indicate standard error.

Table 2, average values (N=5) (in µmol/l) of the total of 5 water samples taken at each, including the standard error (SE)

Forest Intensive Grass Corn

Nutrient Mean (µmol/l) SE Mean(µmol/l) SE Mean(µmol/l) SE

S 259.627 11.39773 305.858 9.938769 183.360 21.72479 P 1.089 0.095836 0.952 0.108057 1.903 0.519076 Fe 6.991 1.031459 1.610 0.033625 6.046 1.575759 Si 104.190 4.946144 116.523 6.539803 109.558 12.88925 Mn 2.196 0.195015 0.255 0.035768 2.717 0.274209 Mg 223.110 9.013539 248.526 7.70401 163.109 18.04429 Ca 652.110 26.48437 643.116 18.1207 465.578 50.12567 Al 8.685 0.203443 11.293 0.326783 7.047 0.741 Na 645.818 19.71056 866.882 26.78642 466.654 52.83371 K 310.425 14.29237 492.756 14.18503 149.447 16.8651 NO3+NO2 278.800 14.15062 638.400 2.767671 97.000 2.50998 NH4 0.800 0.8 0.000 0 4.400 1.208305 Ntot 418.800 15.11754 742.800 7.767883 248.800 5.885576 DON 139.200 4.103657 104.400 6.6 147.400 4.308132 PO4 0.740 0.102956 0.600 0.094868 1.880 0.208327 Cl 334.000 8.115417 479.000 1.854724 240.000 2.437212 SO4 212 9.848858 259 6.4 182 1.989975 pH 6.788 0.034843 6.692 0.016553 6.680 0.023875 sum kath 1181.000 40.46344 1491.000 128.342 843.000 12.52438 sum kath 1212.000 50.59589 1722.000 46.401 873.000 93.22537 Ec 278.000 36.97079 347.000 29.74671 373.000 54.03479 Alkalinity 1122.000 9.557196 941.000 17.64936 1025.000 15.31339

(8)

8 Sediment analysis

Sediment analysis resulted in significant differences in nutrient concentrations between sites. Average values for locations, with standard error, are represented in Appendix 3. C:N ratio (p < .01, F= 21.37), see Figure 5 , and percentage organic matter (OM) (p = 0.020, F=5.52), see Figure 6, differed

Table 3, average nutrient values for the sediment analysis (N=5) Nutr ient For est Intensive Grass Cor n Me an SE Me an SE Me an SE N 0.1 24 0.01 885 0.1 11 0.02 5412 0.0 60 0.01 294 C 1.9 14 0.29 6961 1.4 52 0.33 6354 0.7 96 0.18 5892 P 0.0 31 0.00 4179 0.0 23 0.00 6023 0.0 11 0.00 3728 S 15. 378 0.11 7107 13. 016 0.07 6394 12. 928 0.45 7224 OM 3.2 15 0.46 1844 3.0 62 0.56 7196 1.2 75 0.30 827 C/N 0.0 34 0.00 6187 0.0 49 0.01 7307 0.0 28 0.00 8762 N/P 3.6 81 0.15 7397 3.3 31 1.24 3448 2.7 33 0.81 6756 significantly between treatments. The C:N ratio was highest in forest (mean= 15.38, SE= 0.117) and significantly lower in intensive grass (mean=13.02, SE= 0.076, p <0.01) and Figure 4, DO measured in situ. Intensive Grass was

significantly higher than the rest. Error bars indicate Standard Error

Figuur 5, C:N ratio in the sediment. The ratio was

significantly higher in Forest than in the other locations. Error bars indicate standard error

Figure 7, 28-Day average survival (N=5). Error bars indicate standard error.

Figure 6, Sediment OM content. Only corn was

significantly lower than Forest and Intensive Grass.

Figure 8, 28- Day average growth (N=5) in length (blue bars, mm) and weight (red bars, mg). Error bars indicate standard error.

(9)

9 corn (mean= 12.93, SE= 0.46, p <0.01) site,

while it did not differ significantly between the grass and corn site (p=0.95). OM was high in forest (mean= 3.21%, SE= 0.46) as well as in grass (mean= 3.06, SE= 0.57) and both differed significantly from corn, respectively (p=0.029) and (p=0.043), whereas they did not differ significantly from each other (p=0.97). Corrected for organic matter, no significant differences at all were found.

Bioassays

Survival was generally high but did not differ significantly (p=0.43, Chi2=2.74) between treatments (Figure 7). Mean survival in the laboratory control was 92% (SE=3.74%), mean survival in forest was slightly 94% (SE= 4.0%), mean survival in the intensive grass sample was 86% (SE= 6.78%) and mean survival in corn was 84% (SE=5.10%). Growth in weight and length showed comparable results, but did not differ significantly between locations (see Figure 8). The initial average length of the organisms was 2.68 mm. The average increase in length in the laboratory control was 1.68 mm (SE= 0.139), in forest of 2.31 mm (SE= 0.152), in intensive grass 1.83 mm (SE= 0.164) and in corn 2.01 mm (SE= 0.195). Average weight at the end of the test was 0.188 mg (SE= 0.0239) for the laboratory control, 0.292 mg (SE= 0.0381) for forest, 0.233 mg (SE= 0.0367) for intensive grass and 0.230 mg (SE= 0.0209) for corn. A full list of the survival, length and weight of the organisms at the end of the test can be found in Appendix 2. A regression analysis was performed to assess the relations between the variables survival, weight and length, and the chemicals measured, but none of the chemicals had a significant effect on the response variables tested.

Community composition

General results on the multiple indices used are shown in Table 5. The full list of species abundances per site can be found in Appendix 3. A summary of all species identified is presented in appendix 4. The community composition showed large differences between sites, considering multiple indices. The results of the community composition are shown in Figure 9. In summary, more than twice as many species were present in the forest location compared to the grass as well as to the corn, and the Shannon and Margalef index were also almost twice as high for the forest (Shannon = 2.81, Margalef= 6.86) compared to grass (Shannon = 1.50, Margalef= 3.4) and corn (Shannon = 1.47, Margalef= 3.29), indicating an more equal distribution and higher number of species found per number of individual counted, respectively. The Jaccard Similarity Index also showed more similarity between both anthropogenic impacted landscapes corn and grass ( 42.11%) than between forest and grass (22.41%) or forest and corn (13.04%). Furthermore the EPT index showed a large proportion of sensitive species present in the forest (EPT= 0.17), while there was only one counted in the corn (EPT=0.0021) and grass samples (0.00067), resulting in negligible values. The BBI also assigned the best ecological quality to the forest (BBI= 7), but indicating a slightly better community in the grassland location (BBI= 4) than in the corn (BBI= 3), showing that although both seem quite similar on the basis of the other indices, the first has a slightly more sensitive community present than the latter. The species distribution, based on the taxonomic groups used in determining the BBI and EPT, is shown in Figure 10. The proportion of ologichaetes and chironomids was calculated to express the amount of pollution tolerant species present. This

(10)

10 resulted in the highest value for corn (O/C=

0.64), intermediate values for forest (O/C= Table 3, General results on the community composition. Multiple indices are used. The Jaccard index in the first row is the comparison with Forest and the second row is the comparison between Corn and Intensive Grass.

Index Forest Corn Intensive Grass Total Richness 48 23 25 Shannon 2.72 1.47 1.50 Margalef 6.84 3.29 3.4 Jaccard Forest & 22.41% 13.04% Corn & 42.11% EPT 0.16 0.00067 0.0021 BBI 7 3 4 O Index 0.26 0.64 0.04

0.26) and low values for intensive grass (O/C= 0.04). Sampling method appeared to result in very much differences as well, especially in the forest. The sampling methods were also compared with the Jaccard similarity index. The resemblance between species sampled with the different methods, surber and the handnet, was only 20.9% for the forest, 57.9% for the corn and 72.7 for the intensive grass. Discussion

The chemical water analysis showed significant differences in macro nutrients nitrogen, sulphur and phosphate, and micro

nutrient (Fe, Mg, Mn, Ca, Na, K, Al, NO3, Cl and SO4) composition, indicating a land use specific composition of nutrients, (this highlights the probability that other concentrations are also land use specific) which makes it also likely for other pollutants such as herbicides, antibiotics and metals. Total nitrogen differed significantly between sites and exceeded environmental quality

Figure 10, Abundances of the major taxonomic groups present in different locations. The same taxonomic groups are used as for the BBI and EPT. Differences are evident for the sensitive EPT

(11)

11 standards for good water quality in all

locations, (STOWA 2013). It did however, not significantly affect survival. The same result was found by Friberg et al. (2010). They showed that nitrogen pollution did not have much effect on community composition and indicate that phosphorous (which was not significantly different in the water samples), and nitrate have more detrimental effects on community composition. showed that

nitrogen has fewest effects on benthic survival and community composition of benthic invertebrates whereas nitrate and phosphorous are more important and detrimental for community composition and survival. Nitrate water levels exceeded environmental levels for natural systems, and even reached eutrophic levels (Smith et al., , although there was no significant effect on survival. There were also differences in sediment nutrient composition, although there are no standards to compare them with. Food quality (C:N ratio) and quantity (organic matter) differed between sites, resulting in the highest C:N ratio in the forest site and lower ratios in the grassland and corn sites. Organic matter was highest in the forest and corn. Changes in food quality and quantity can be detrimental for community composition (Strand & Merrit, 1999).

High nutrient levels and increased algal growth may result in decreased oxygen levels when algae start decaying. Based on the high nutrient levels, it is possible that algae growth is simulated in summer, resulting in depleted oxygen levels. Research from Friedberg et al. (2010) and Daueret al., (2000)showed a strong negative correlation between low oxygen levels and macro-invertebrate community composition. At time of measurement, the oxygen saturation was above 90% for all locations. This however, may change with seasons. Light, warmth and slow stream velocity which are occurring in summer, act all

stimulate the decaying process synergistically resulting in depleted oxygen levels, and macro invertebrate mortality. This effect is difficult to account for since good aeration and stable temperatures will compensate for these effects in the bioassays.

Both survival and growth were highest in forest, although it was not significant. Regression analysis resulted in no significant effect of the single nutrients on survival or growth. These results do indicate the absence of toxicity in the sample for Hyalella azteca, and since the Hyalella azteca is proven to be a sensitive organism (Borgmann et al., 1989), it is possible that they were not present at the time of sampling. Although no significant results were found, a longer test taking reproduction into account and multiple trophic levels might represent more population dynamics, which is illustrated by (Peter M Chapman et al., 1997; Postma & Davids, 1995). The high survival rates indicated good water and probably also food quality in all treatments.

Based on the community composition, there are major differences between the reference location, forest, and the locations related to agriculture. Most indices used give higher values for forest, except for the OC index and the Jaccard Similarity Index. The Jaccard similarity between the two sampling methods resulted similarities which were sometimes lower within than between sites. It is likely that this effect is cause due to habitat heterogeneity, since this was also highest in forest whereas it was low for corn and intensive grass, were the similarity between methods was also much higher. These results are indicating that this is of major importance for the community composition, as is supported by Allan (2011). Species diversity is higher and more equally distributed in the than in the grass and corn samples. For sensitive species, the EPT index it is negligible

(12)

12 for the intensive grass and corn samples

whereas it is very high for the forest sample. Furthermore the Belgian Biotic index indicated good water quality in the forest (BBI= 7) whereas it indicated deteriorated water quality in the grass and corn samples. At last the proportion pollution tolerant species, oligichates and chironomids, was also calculated, the same way as the EPT, proportion of sensitive species, was calculated. This resulted in the lowest value in the forest sample (OC= 0.45), a slightly higher value in the corn sample (OC= 0.49) and the highest value in the grass sample (OC= 0.68). The high abundance of oligochaetes and chironomids supports the hypothesis that low oxygen levels are causing macro invertebrate mortality, since these organisms are known to be more resistant for low oxygen levels(Allan, 2011; Friberg et al., 2010).

A possible explanation for the contradicting results found in the ecology and the chemistry and toxicology is the timing. Concentrations fluctuate during the year and depending on the frequency of activity. Therefore the results found do not have to be representative for the full year. This can bias the results from the chemical analysis and bioassays, but the community composition is the most time integrated test, representing the effect regardless of daily fluctuations and accounting for effects on population level. Chemical make-up of the streams may differ from day to day, dependent on weather circumstances and farming activities. The values found in the chemical analysis may therefore not be representative for the whole year. The time frame used for the chemical analysis also is important for the bioassay, since this the same sediment and water samples are used. The bioassay represents slightly longer effects than the chemical analysis, although a full 48-day bioassay taking reproduction into account might be even

more sensitive and relevant, since population dynamics might be altered (Postma & Davids, 1995). The last part of the triad, which is the community assembly, is the most time integrated one and represents years of effects.

It can be questioned however, if the list of compounds was conclusive enough, since the majority of compounds measured were nutrients, there were no heavy metals, and land use specific compounds such as herbicides, antibiotics and other abiotic pollutants measured. Absence of data on these pollutants does not mean that there is none. Given these results, it remains a question why there are effects on the community level, while there were no effects in the toxicity test or the chemical analysis. Combining all three triad parts, chemistry, toxicity and ecology, should together provide multiple lines of evidence of pollution-induced degradation. The data are compared to reference site values, in this case the forest location, which is the least contaminated site and is indicative for (regional) background values, according to the method described by Chapman (1990). Nutrients in water and sediment differed significantly between sites. These compounds can be used as indicator compounds representing a different chemical composition. For the bioassays no significant

Figure 11, Visual presentation of differences found compared to the reference location forest.

(13)

13 difference in outcome was reported. The

ecology however showed severe effects of land use on benthic invertebrates, which indicates that there are differences between sites. Since the community composition is the most time integrated one it is of major importance. The community composition is the most time integrated part from the TRIAD, since it is the result of multiple years of effect. On the basis of these results, according to the conclusions from Chapman (1990), one can say that effects are evident and land use specific, but not due to compounds measured. However land use specific effects are evident and the significant differences between chemical composition can be used as an indicator for unmeasured compounds. Including other chemicals to the list of measured compounds in future research might cause the results of the chemical analysis to a more profound indication of pollution induced degradation.

Acknowledgments

This study was made possible by the department of aquatic biology from the UvA, and it would not have been possible without the advice of my supervisors P.C. Dos Reis Oliveira and M.H.S. Kraak, and the help of my co-workers T. Theirlynck and J. Pasqualini.

(14)

14 References

Allan, J. D. (2011). Landscapes and

Riverscapes : The Influence of Land Use on Stream Ecosystems. Annual Review of Ecology , Evolution , and Systematics, 35(2004), 257–284.

Allan, J. D. (2011). Landscapes and

Riverscapes : The Influence of Land Use on Stream Ecosystems Author ( s ): J . David Allan Source : Annual Review of Ecology , Evolution , and Systematics , Vol . 35 ( 2004 ), pp . 257-284 Published by : Annual Reviews Stable URL : http://w. Annual Review of Ecology , Evolution , and Systematics, 35(2004), 257–284.

Artlett, A. D. J. B., Alakrishnan, V. K. B., Oito, J. T., & Rown, L. R. B. (2013). Toxicity of four sulfonamide antibiotics to the freshwater aphipod Hyalella azteca. Environmental Toxicology and Chemistry, 32(4), 866–875.

http://doi.org/10.1002/etc.2129

Borgmann, U., Grapentine, L., Norwood, W. P., Bird, G., Dixon, D. G., Lindeman, D., … Wang, F. (2005). Sediment toxicity testing with the freshwater amphipod Hyalella azteca: Relevance and

application. Chemosphere, 61(11), 1740– 1745.

http://doi.org/10.1016/j.chemosphere.2 005.04.011

Borgmann, U., Ralph, K. M., & Norwood, W. P. (1989). Toxicity test procedures for Hyalella azteca, and chronic toxicity of cadmium and pentachlorophenol to H-azteca, Gammarus-fasciatus, and Daphnia magna. Archives of Environmental Contamination and Toxicology, 18, 756–764. Retrieved from files

CBS. (n.d.). CBS StatLine - Bodemgebruik; uitgebreide gebruiksvorm, per gemeente. Retrieved April 20, 2016, from

http://statline.cbs.nl/StatWeb/publicatio n/default.aspx?DM=SLNL&PA=70262NED

&D1=0,24&D2=0&D3=a&HDR=T&STB=G 1,G2&VW=T

Chapman, P. M. (1990). The sediment quality triad approach to determining pollution-induced degradation. Science of the Total Environment, The, 97-98(C), 815–825.

http://doi.org/10.1016/0048-9697(90)90277-2

Chapman, P. M. (1996). Presentation and interpretation of Sediment Quality Triad data. Ecotoxicology (London, England), 5(5), 327–39.

http://doi.org/10.1007/BF00119054 Chapman, P. M., Anderson, B., & Carr, S.

(1997). General guidelines for using the sediment quality triad. Marine Pollution …, 34(6), 368–372. Retrieved from http://scholar.google.com/scholar?hl=en &btnG=Search&q=intitle:General+Guidel ines+for+using+the+Sediment+Quality+T riad#0 Cincinnati, O. (2000). EPA/800/R-99/064. Washington DC.

Dauer, D. M., Weisberg, S. B., & Ranasinghe, J. A. (2000). Relationships between Benthic Community Condition, Water Quality, Sediment Quality, Nutrient Loads, and Land Use Patterns in Chesapeake Bay. Estuaries, 23(1), 80.

http://doi.org/10.2307/1353227 Daughton, C. G., & Ternes, T. A. (1999).

Pharmaceuticals and personal care products in the environment: agents of subtle change? Environmental Health Perspectives, 107 Suppl , 907–38. Retrieved from

http://www.pubmedcentral.nih.gov/arti clerender.fcgi?artid=1566206&tool=pmc entrez&rendertype=abstract

De la Torre, A. I., Jiménez, J. A., Carballo, M., Fernandez, C., Roset, J., & Muoz, M. J. (2000). Ecotoxicological evaluation of pig slurry. Chemosphere, 41(10), 1629–1635.

http://doi.org/10.1016/S0045-6535(00)00038-2

(15)

15 Ecotoxicology of human

pharmaceuticals. Aquatic Toxicology (Amsterdam, Netherlands), 76(2), 122– 59.

http://doi.org/10.1016/j.aquatox.2005.0 9.009

Friberg, N., Skriver, J., Larsen, S. E., Pedersen, M. L., & Buffagni, A. (2010). Stream Macroinvertebrate Occurrence Along Gradients in Organic Pollution and Eutrophication. Freshwater Biology, 55(7), 1405–1419.

http://doi.org/10.1111/j.1365-2427.2008.02164.x

Halling-Sørensen, B., Nors Nielsen, S., Lanzky, P. F., Ingerslev, F., Holten Lützhøft, H. C., & Jørgensen, S. E. (1998). Occurrence, fate and effects of pharmaceutical substances in the environment- A review. Chemosphere, 36(2), 357–393.

http://doi.org/10.1016/S0045-6535(97)00354-8

In, I. (2013). Groundwater Assessment and Protection Fact Sheet :

Ingersoll, C. G., Gerald, T. A., Duane, A. B., Eric, L. B., Allen, G. B., James, F. D., … Parley, V. W. (1995). Tocicity and

Bioaccumulation of Sediment-Associated Contaminants using Freshwater

Invertebrates: A Review of Methods and Applications. Environmental Toxicology, 14(11), 1885–1894.

Ingersoll, C. G., Kunz, J. L., Hughes, J. P., Wang, N., Ireland, D. S., Mount, D. R., … Valenti, T. W. (2015). Relative sensitivity of an amphipod Hyalella azteca, a midge Chironomus dilutus, and a unionid mussel Lampsilis siliquoidea to a toxic sediment. Environmental Toxicology and Chemistry / SETAC, 34(5), 1134–44. http://doi.org/10.1002/etc.2909 Leslie, H. A., Pavluk, T. I., Vaate, A. bij de, &

Kraak, M. H. S. (1999). Triad Assessment of the Impact of Chromium

Contamination on Benthic

Macroinvertebrates in the Chusovaya River (Urals, Russia). Archives of

Environmental Contamination and Toxicology, 37(2), 182–189.

http://doi.org/10.1007/s002449900504 Muñoz, I., López-Doval, J. C., Ricart, M.,

Villagrasa, M., Brix, R., Geiszinger, A., … Barceló, D. (2009). Bridging levels of pharmaceuticals in river water with biological community structure in the Llobregat River basin (northeast Spain). Environmental Toxicology and Chemistry / SETAC, 28(12), 2706–14.

http://doi.org/10.1897/08-486.1 OECD. (2004). Guideline 218: Sediment-water

Chironomid toxicity test using spiked sediment. Parice, France.

Pauw, de, N., & Vannevel, R. (1993). Macro-invertebraten en waterkwaliteit. Stichting Leefmilieu, 175–266.

Plumb, R. H. (1981). Procedures for handling and chemical analysis of sediment and water samples. United States

Environmental Protection Agency/Corps of Engineers Technical Committee on Criteria for Dreged and Fill Material, EPA/CE-81-, 507. Retrieved from

http://oai.dtic.mil/oai/oai?verb=getReco rd&metadataPrefix=html&identifier=AD A103788

Postma, J. F., & Davids, C. (1995). Tolerance Induction and Life Cycle Changes in Cadmium-Exposed Chironomus riparius (Diptera) during Consecutive

Generations. Ecotoxicology and Environmental Safety, 30(2), 195–202. http://doi.org/10.1006/eesa.1995.1024 R Core Team. (2014). R: A language and

environment for statistical computing. Vienna: R Foundation for Statistical Compution. Retrieved from

http://www.r-project.org/

Smith, A. J., Bode, R. W., & Kleppel, G. S. (2007). A nutrient biotic index ( NBI ) for use with benthic macroinvertebrate communities, 7, 371–386.

http://doi.org/10.1016/j.ecolind.2006.03 .001

(16)

16 Strand, M., & Merrit, W. R. (1999). Impacts of

livestock grazing activities on stream insect communities and the riverine environment. American Entomologist, 45(1), 13–29.

Strayer, D. L., Beighley, R. E., Thompson, L. C., Brooks, S., Nilsson, C., Pinay, G., & Naiman, R. J. (2003). Effects of land cover on stream ecosystems: Roles of empirical models and scaling issues. Ecosystems, 6(5), 407–423.

http://doi.org/10.1007/s10021-002-0170-0

Verdonschot, P. F. M., & Nijboer, R. C. (2002). Towards a decision support system for stream restoration in the Netherlands : an overview of restoration projects and future needs, (1986), 131–148.

Wagenhoff, A., Townsend, C. R., & Matthaei, C. D. (2012). Macroinvertebrate

responses along broad stressor gradients of deposited fine sediment and dissolved nutrients : a stream mesocosm

experiment, 892–902.

http://doi.org/10.1111/j.1365-2664.2012.02162.x

Wolfram, G., Höss, S., Orendt, C., Schmitt, C., Adámek, Z., Bandow, N., … de Deckere, E. (2012a). Assessing the impact of chemical pollution on benthic invertebrates from three different European rivers using a weight-of-evidence approach. Science of the Total Environment, 438(2012), 498–509. http://doi.org/10.1016/j.scitotenv.2012. 07.065

Wolfram, G., Höss, S., Orendt, C., Schmitt, C., Adámek, Z., Bandow, N., … de Deckere, E. (2012b). Assessing the impact of chemical pollution on benthic invertebrates from three different European rivers using a weight-of-evidence approach. The Science of the Total Environment, 438, 498–509. http://doi.org/10.1016/j.scitotenv.2012. 07.065

(17)

17 nr# Si te S m ol/ L ) P m ol/ L ) Fe m ol/ L ) Si m ol/ L ) Mn m ol/ L ) Mg m ol/ L ) Ca m ol/ L ) Al m ol/ L ) Na m ol/ L ) K m ol/ L ) N O 3+ N O 2 [u m ol/ l N H4 [u m ol/ l ] N to t [u m ol/ l ] DO N [u m ol/ l ] PO 4 [u m ol/ l ] Cl [u m ol/ l ] SO 4 [u m ol/ l ] pH Alk alin ity su m ka th eq /L ) su m ka th eq /L ) Ec S/ cm ) 1 Fo re st 275 1. 30 4 6. 99 11 0. 7 2. 69 241 696 9. 22 681 334 287 0 434 147 0.5 333 243 6.6 6 11 53 .0 1257 1270 287 2 Fo re st 215 0.7 85 8. 11 86 .6 2. 20 189 550 8. 44 573 256 223 4 359 132 1.1 303 181 6.7 7 97 2.5 984 1073 240 3 Fo re st 274 1. 28 6 6. 92 11 5. 8 2. 26 233 691 9. 01 678 332 288 0 435 147 0.7 342 215 6.8 4 11 10 .5 1254 1269 287 4 Fo re st 266 1. 02 2 6. 68 10 4. 3 2. 34 227 666 8. 70 639 315 299 0 426 127 0.6 350 209 6.8 5 11 68 .5 1204 1200 288 5 Fo re st 269 1. 05 0 6. 25 10 3. 5 1. 49 226 656 8. 07 657 315 297 0 440 143 0.8 340 212 6.8 2 12 06 .5 1208 1249 289 6 I.G ra ss 281 0.6 50 1. 55 11 8. 4 0. 27 234 599 10 .77 793 468 639 0 756 117 0.7 475 240 6.7 0 93 1.5 1399 1674 351 7 I.G ra ss 319 1. 19 2 1. 72 12 5. 0 0. 33 254 664 11 .93 886 506 639 0 756 117 0.6 482 255 6.7 2 96 5.0 1543 1767 350 8 I.G ra ss 283 0. 75 4 1. 54 93 .1 0. 26 229 600 10 .33 815 451 628 0 718 90 0.4 477 252 6.6 4 91 9.0 1360 1630 330 9 I.G ra ss 316 1. 15 6 1. 64 13 1. 6 0. 29 259 674 12 .00 927 519 644 0 731 87 0.9 477 272 6.7 3 94 0.0 1566 1754 352 10 I.G ra ss 330 1. 00 9 1. 60 11 4. 5 0. 12 266 679 11 .43 913 521 642 0 753 111 0.4 485 274 6.6 7 94 8.0 1587 1785 353 11 E.G ra ss 345 6. 44 5 12 .07 75 .2 4. 64 271 655 10 .51 783 745 316 129 593 148 5.5 439 328 6.8 2 15 95 .0 1846 1542 21 0.3 12 E.G ra ss 375 7. 11 0 13 .42 79 .6 5. 14 301 727 11 .47 816 808 304 134 592 154 5.6 435 300 6.7 9 16 20 .0 2010 1580 21 3.7 13 E.G ra ss 337 6. 32 7 12 .16 70 .6 4. 31 264 637 10 .19 744 730 298 134 597 165 5.1 427 308 6.7 8 15 90 .0 1806 1523 21 3.7 14 E.G ra ss 349 6. 44 3 12 .17 71 .5 4. 85 262 665 10 .67 765 738 311 133 605 161 5.9 425 323 6.7 6 15 40 .0 1839 1549 21 3.6 9 15 E.G ra ss 360 7. 23 5 12 .37 72 .3 3. 77 276 678 10 .44 777 765 310 129 584 145 5.8 406 356 6.7 5 14 60 .0 1907 1524 21 4.2 16 Co rn 206 3. 09 7 11 .84 12 1. 7 3. 24 175 501 7. 29 517 172 87 7 231 137 2.2 240 187 6.6 0 99 9.5 945 896 378 17 Co rn 204 2. 07 5 6. 03 12 0. 5 2. 86 182 516 7. 75 517 165 99 6 258 153 1.8 247 181 6.7 5 10 20 .5 928 938 375 18 Co rn 198 1. 88 4 5. 18 11 6. 5 2. 67 175 491 7. 68 495 154 99 5 265 161 1.5 235 184 6.6 9 10 30 .0 894 930 371 19 Co rn 213 2. 46 1 4. 88 13 0. 3 3. 11 192 551 8. 35 547 173 100 4 245 141 1.4 234 175 6.6 8 10 03 .5 976 942 369 20 Co rn 97 0. 00 0 2. 30 58 .8 1. 69 92 269 4. 16 258 83 100 0 245 145 2.5 243 181 6.6 8 10 69 .5 474 660 370

(18)

18 Appendix 2, Full list of chemicals measured in the sediment and p-values for OM

Tabel 1 Nutrient concentrations in the sediment, shown in percentage /kg sediment

Tabel 2 Nutrient concentrations, %/kg, corrected for organic matter

ID Location N.OM C.OM S.OM P.OM

1 Forest 0.042989 0.642864 0.011044 0.010755 2 Forest 0.034096 0.526031 0.01006 0.009259 3 Forest 0.038051 0.584097 0.009777 0.0113 4 Forest 0.038934 0.6047 0.007923 0.009581 5 Forest 0.038587 0.602294 0.009175 0.011719 6 Intensive Grass 0.018489 0.236822 0.006648 0.008673 7 Intensive Grass 0.036802 0.475272 0.00429 0.011336 8 Intensive Grass 0.04398 0.577136 0.007695 0.058127 9 Intensive Grass 0.046588 0.619712 0.009378 0.018748 10 Intensive Grass 0.034243 0.443521 0.008131 0.004261 11 Corn 0.076302 1.055152 0.010639 0.013359 12 Corn 0.038845 0.46614 0.008879 0.012785 13 Corn 0.069231 0.971426 0.010262 0.030175 14 Corn 0.03014 0.345726 0.003723 0.028652 15 Corn 0.038349 0.506192 0.010326 0.024457 Replicate Location N [%] C [%] S [%] C/N ratio OM (%) P (%) N/P 1 Forest 0.1304 1.95 0.0335 14.96 3.033303 0.032624 3.997008 2 Forest 0.0888 1.37 0.0262 15.45 2.604411 0.024114 3.682543 3 Forest 0.0899 1.38 0.0231 15.31 2.362621 0.026698 3.367262 4 Forest 0.1204 1.87 0.0245 15.54 3.09244 0.029628 4.063749 5 Forest 0.1922 3 0.0457 15.63 4.980954 0.058369 3.29282 1 Intensive Grass 0.0445 0.57 0.016 12.89 2.406875 0.020874 2.131805 2 Intensive Grass 0.0875 1.13 0.0102 12.87 2.377587 0.026953 3.246441 3 Intensive Grass 0.0823 1.08 0.0144 13.09 1.871311 0.108773 0.756621 4 Intensive Grass 0.1699 2.26 0.0342 13.28 3.646856 0.068371 2.484987 5 Intensive Grass 0.1714 2.22 0.0407 12.95 5.005406 0.02133 8.035772 1 Corn 0.0875 1.21 0.0122 13.79 1.146754 0.015319 5.711755 2 Corn 0.035 0.42 0.008 12.11 0.901017 0.01152 3.038264 3 Corn 0.0506 0.71 0.0075 13.92 0.730884 0.022055 2.294315 4 Corn 0.034 0.39 0.0042 11.62 1.128061 0.032321 1.051945

(19)

19 Appendix 4

P-values of the statisical analysis of the sediment nutrient values. Nutrient Test ANOVA Kruskal-Wallis N 0.677 C 0.330 P 0.069 S 0.145 OM 0.020* C/N <0.000* N/P 0.741

(20)

20

Forest Intensive Grassland Corn

Anabolia nervosa 16 0 0 Androposopa sp 2 0 0 Asellus aquaticus 2 0 0 Asellus sp 5 1 206 Athripsodes cinereus 26 0 0 Athripsodes sp 5 0 0 Baetis vernus 3 1 0 Bathyomphalus contortus 2 0 0 Ceratopogonidae sp 1 0 0 Chaoboridae sp 1 0 0 Chironomidae 0 2 1 Chironominae sp 0 0 0 Chironomini sp 16 0 0 Chironomus sp 22 196 26 Chrysops sp 26 0 0 Corixa sp 1 0 0 Culicoidini sp 26 1 0 Dytiscus sp 0 5 0 Enchytraeidae sp 0 0 1 Ephemera danica 50 0 0 Ephemera sp 2 0 0 Erpobdella octoculata 2 0 0 Galba truncatula 1 12 0 Gammarus pollux 149 9 0 Glossiphonia complanata 10 0 0 Gyrinus sp 11 1 0 Halesus radiatus 1 0 0 Haliplus sp 2 0 0 Helophilus pendulus 1 0 0 Helophorus brevipalpis 0 0 5 Hydrobius fuscipes 0 0 2 Hydrobius sp 0 0 1 Hydrophilus caraboides 0 0 1 Hydroporus palustris 0 0 1 Laccobius sp 0 0 3 Leptoceridae sp 2 0 0

(21)

21 Limnephilus lunatus 23 0 1 Limonidae pilaria 0 0 1 Limonidae sp 1 0 0 Lithoglyphos naticoidos 0 2 0 Megasternum obscurum 0 0 1 Mesophylax punctatus 2 0 0 Micropterna lateralis 8 0 0 Myxas glutinosa 1 12 0 Naidae sp 215 0 0 Naididae (hairy) 0 3 0 Naididae sp 0 11 0

Naididae sp 3 (blue stripe) 5 0 0

Naididea sp 2 (hairy) 2 0 0 Nepa cinera 0 0 1 Noterus sp 7 0 0 Notidobia ciliaris 2 0 0 Oligochaeta sp 0 0 18 Orthocladiinae sp 14 6 10 Paraphaenocladius sp 3 0 0 Pedicia crunobia sp 0 16 26 Physa fontinalis 0 136 0 Physella acuta 0 1 0 Pisidium Casertanum 17 78 0 Platambus maculatus 0 5 16 Proasellus meridianus 4 0 193 Ptychopteridae sp 0 3 17 Spercheus emarginatus 0 0 1 Stylaria lacustris 1 5 870 Tanypodinae sp 43 9 9 Tanytarsini sp 81 0 6

(22)

22

(23)

Referenties

GERELATEERDE DOCUMENTEN

The functional group the conditioning plant species belonged to, grass or forb, significantly explained the distribution of plant biomass over the plant species during the

Lasse Lindekilde, Stefan Malthaner, and Francis O’Connor, “Embedded and Peripheral: Rela- tional Patterns of Lone Actor Radicalization” (Forthcoming); Stefan Malthaner et al.,

Muslims are less frequent users of contraception and the report reiterates what researchers and activists have known for a long time: there exists a longstanding suspicion of

Coefficients that have zero value under statistical in- dependence, maximum value unity, and minimum value minus unity independent of the mar- ginal distributions, are the

For aided recall we found the same results, except that for this form of recall audio-only brand exposure was not found to be a significantly stronger determinant than

managers offering their services to clients with holdings under $500.000,- are obligated to..  In many other countries like the Netherlands, Italy etc. regulation is less tight

Die probleme wat deur damesuitrusters ondervind word, word.

Binne die gr·oter raamwerk van mondelinge letterkunde kan mondelinge prosa as n genre wat baie dinamies realiseer erken word.. bestaan, dinamies bygedra het, en