Chronic nutrient inputs affect stream macroinvertebrate communities more than acute inputs: An experiment manipulating phosphorus, nitrogen and sediment

Highlights

• Addressed effects of chronic vs. acute stressors which are poorly understood

• Sediment, phosphorus and nitrogen manipulated

• Chronic nutrients and sediment were added for 18 days.

• Two three-hour pulses of acute nutrient (concentrations doubled)

• Responses were generally as a result of chronic exposure and sediment was the most pervasive stressor.

Abstract

Freshwaters worldwide are affected by multiple stressors. Timing of inputs and pathways of delivery can influence the impact stressors have on freshwater communities. In particular, effects of point versus diffuse nutrient inputs on stream macroinvertebrates are poorly understood. Point-source inputs tend to pose a chronic problem, whereas diffuse inputs tend to be acute with short concentration spikes. We manipulated three key agricultural stressors, phosphorus (ambient, chronic, acute), nitrogen (ambient, chronic, acute) and fine sediment (ambient, high), in 112 stream mesocosms (26 days colonisation, 18 days of manipulations) and determined the individual and combined effects of these stressors on stream macroinvertebrate communities (benthos and drift). Chronic nutrient treatments continuously received high concentrations of P and/or N. Acute channels received the same continuous enrichment, but concentrations were doubled during two 3-hour periods (day 6, day 13) to simulate acute nutrient inputs during rainstorms. Sediment was the most pervasive stressor in the benthos, reducing total macroinvertebrate abundance and richness, EPT (mayflies, stoneflies, caddisflies) abundance and richness. By contrast, N or P enrichment did not affect any of the six studied community-level metrics. In the drift assemblage, enrichment effects became more prevalent the longer the experiment went on. Sediment was the dominant driver of drift responses at the beginning of the experiment. After the first acute nutrient pulse, sediment remained the most influential stressor but its effects started to fade. After the second pulse, N became the dominant stressor. In general, impacts of either N or P on the drift were due to chronic exposure, with acute nutrient pulses having no additional effects. Overall, our findings imply that cost-effective management should focus on mitigating sediment inputs first and tackle chronic nutrient inputs second. Freshwater managers should also take into account the length of exposure to high nutrient concentrations, rather than merely the concentrations themselves.

Introduction

Freshwaters are among the world's most endangered ecosystems (Sala et al., 2000) and river degradation and the loss of biodiversity are of major concern worldwide (Strayer and Dudgeon, 2010; Vörösmarty et al., 2010), with activities such as forestry, urbanisation and agriculture detrimentally affecting the integrity of aquatic ecosystems (Strayer and Dudgeon, 2010). Most river catchments are affected to some degree by agricultural activities (Allan, 2004), which leads to impacts from nutrient enrichment, principally nitrogen (N) and phosphorus (P), and increased fine sediment deposition (Allan, 2004; Matthaei et al., 2010). The effects of stressors acting on freshwater ecosystems has received much attention in recent years (Beermann et al., 2018; Graeber et al., 2017; Lemm and Feld, 2017; Ormerod et al., 2010). The emerging consensus is that there are a multitude of influential stressors, which often interact in complex and unexpected ways (Jackson et al., 2016; Nõges et al., 2016). Consequently, there is a growing recognition that an understanding of how multiple stressors interact and the implications these interactions have for ecological communities are required for effective resource management.

Deposited fine sediment has been shown to affect stream ecosystems by infilling of open gravel substrata (Soulsby et al., 2001), leading to declines in habitat quality (Wood and Armitage, 1997), which in turn causes changes in benthic community composition and increases the propensity of stream macroinvertebrates to drift (Larsen and Ormerod, 2010; Rosenberg and Wiens, 1978). Fine sediment often has the strongest detrimental effects on pollution-sensitive groups including Ephemeroptera, Plecoptera and Trichoptera (EPT) (Jones et al., 2012; Larsen et al., 2009; Matthaei et al., 2006; Rabení et al., 2005; Townsend et al., 2008). Richness and abundances of these groups tend to decline in response to sediment addition (Jones et al., 2012), while in some cases positive effects have been found for pollution-tolerant groups such as Chironomidae (Kreutzweiser et al., 2005).

The two main origins of pollution in freshwaters are point sources (e.g. waste water effluent) and diffuse sources (e.g. organic and inorganic nutrients in runoff). Diffuse pollution is usually associated with large, short-lived spikes (typically 3–6 h) in nutrient and sediment concentrations during storm events (Bowes et al., 2008; Jordan et al., 2007; Jordan et al., 2012). This type of pollution is referred to here as ‘acute’. By contrast, point sources pose a more chronic threat, having the greatest effect during the summer months when water levels are generally low (Jarvie et al., 2006; Withers et al., 2014), and are referred to here as ‘chronic’.

As observed elsewhere in Europe (OECD, 2012), unsatisfactory water quality in Ireland has been attributed to diffuse agricultural pollution and point discharges of waste water (Fanning et al., 2017). Acute pollution events, principally derived from agriculture, often comprise the majority of annual pollutant loads in rural catchments and are typically associated with storm events (Smith et al., 2005). However, chronic point-sources are often sufficient to maintain rivers in a eutrophic state between storm events (Jordan et al., 2007). It has also been suggested that chronic nutrient inputs may pose a greater risk to river water quality than acute storm inputs due to the ecological risks associated with symptoms of eutrophication, such as excessive algal growth in times of ecological sensitivity such as summer low-flows, when biological activity is high (Jarvie et al., 2006). Thus, tackling chronic inputs of nutrients during low summer flows may provide the greatest gains in terms of improving ecological quality (Shore et al., 2017).

The question remains as to the impact of an acute (diffuse) event, if it also occurs during a period of ecological sensitivity (e.g. summer low flows) when chronic (point) pressures usually dominate. Further, it is projected that due to climate change, storms will increase in frequency and intensity throughout the year, resulting in an increased risk of diffuse pollution (Easterling et al., 2000). It is still unclear what impact this may have on ecosystems which are already under pressure from chronic pollution sources at these times, but it is predicted that increased frequency of extreme events will have serious consequences for ecosystem food webs (Woodward et al., 2016). In addition, little work has been carried out on the impact of chronic versus acute nutrient pressures on ecological communities and this work has focussed on single stressors, most notably phosphorus (e.g. Shore et al., 2017). However, agricultural stressors rarely occur on their own (Ormerod et al., 2010; Townsend et al., 2008), therefore efforts to remediate their effects must reflect a holistic approach which takes into account the impact of multiple stressors acting simultaneously. Thus, there is a need for experimental studies assessing the relative influence of chronic and acute sources of nutrient pollution in a multiple stressor context.

To address the knowledge gaps identified above, a mesocosm experiment was undertaken to investigate the response of stream macroinvertebrate communities to nitrogen, phosphorus and sediment, with the two nutrients being manipulated to simulate the effects of both chronic and acute pollution events, with acute events occurring against a backdrop of chronic pollution. This study builds on a related study (Davis et al., 2018) which examined the effects of the same three stressors by distinguishing between chronic and acute effects of the two nutrients. Based on findings from this previous study and those of related experiments in New Zealand (Piggott et al., 2015; Wagenhoff et al., 2012) and Germany (Elbrecht et al., 2016), it was hypothesised that 1) sediment would be the most pervasive stressor, consistently having a negative effect on sensitive macroinvertebrate taxa in the benthos, 2) sediment addition would lead to an immediate strong drift response but this would diminish as time went on, 3) nutrients would have relatively weak (compared to sediment effects) but predominantly negative effects on sensitive macroinvertebrates in the benthos, 4) nutrients would cause no immediate drift response but such responses would develop with longer exposure, 5) there would be a greater negative effect of acute nutrient enrichment on macroinvertebrates, beyond that seen in channels which only received chronic nutrient enrichment, and 6) the most severe impact would be seen in mesocosms which received sediment coupled with an acute nutrient treatment. To the best of our knowledge, this is the first experiment to investigate the effects of chronic versus acute nutrient enrichment with N and P on stream macroinvertebrate communities.