Monitoring of an urban lake in the Mediterranean

Urban lakes are artificial systems that accomplish many functions, such as storing rainwater, avoiding flooding 10 of adjacent urban areas and supporting recreational activities. However, their intrinsic aesthetic value is usually reduced 11 due to eutrophication problems and anoxia processes. The objective of this study is to present the results of the water 12 quality monitoring of a small urban lake (11264 m2 and 1.5 m average depth) in Tavernes de la Valldigna (Valencia, Spain) 13 during summer 2016. The final aim is to determine the better parameters for monitoring urban lakes having into account 14 budget restrictions. La Goleta lake has suffered repeated events of fish deaths and bad odors that cause the alarm of 15 residents and tourists, especially in summer. Municipal authorities undertook a restoration project which first part was 16 developed during the first semester of 2016. Surveillance monitoring should be financed by the Town Council, so limiting 17 the monitored parameters to the most appropriate ones is key for guarantying long-term surveillance. The results of this 18 study show the importance of macrophyte community in determining water quality and maintaining dissolved oxygen 19 levels. Dissolved oxygen is a key parameter easy to measure and a good indicator of lake water quality evolution. 20 Analytical methodologies must be adapted to the high organic matter content of these systems to avoid interferences. 21


25
Urban lakes have been described as smaller and shallower water bodies than natural lakes, with 26 a larger ratio of watershed area to lake surface area [1]. This causes a greater exposition of urban lakes 27 to human impacts. Eutrophication issues have been well studied in natural lakes and the effects of 28 harmful algal blooms (HABs) have become a growing concern for water resources management.

29
However, studies focussing on urban lakes are rare and scientits have pointed out the need of a

31
Cities benefit from internal urban ecosystems, such as urban lakes thanks to the ecosystem 32 services that they offer [2]. Direct ecosystems services of urban lakes can be rainwater drainage, 33 storing rainwater, water supply, recreational and cultural values. The locally generated ecosystem 34 services have a substantial impact on the quality-of-life in urban areas [2]. To preserve these services 35 we need to be sure that management strategies are effective, thus, we need effective monitoring 36 programs able to detect relevant water quality changes.

37
Study area 38 La Goleta Lake is an urban lake located at Tavernes de la Valldigna town (Eastern Spain) ( Figure   39 1). It works as storm tank that collects runoff of nearly 200,000 m 2 of urban area. Its current 40 dimensions are 11,264 m 2 and 1.5 m average depth. For a complete description see [3]. This town is a 41 very important touristic destination in the Spanish Mediterranean area. In fact tourism is one of the was developed during the first semester of 2016. A closed circuit for recirculating water was built 48 with element such as fountains and waterfalls to increase water aeration. Also, UV clarifiers were 49 coupled to the recirculation system. For more details on these restoration measures see [3].

50
During construction of the recirculation system, the water level of the lake was lowered by 51 pumping water to the sea. The penetration of sunlight to the bottom of the lake allowed the 52 development of a benthic substrate dominated by the green algae Chara sp. It has been described that 53 Characea occurs in shallow parts of lakes (0.5-2.5m), provided that water quality has sufficiently 54 improved and enough light penetrates to the soil [5]. The observed recolonisation process was only

70
La Goleta lake was usually sampled at the same hour between 9:00 and 10:00h a.m. The sampling 71 frequency was biweekly, but was increased after precipitation events because of the well-known

107
Gonyaulax sp. and Gymnodinium sp. and to centric diatoms such a Navicula sp., Amphora sp. and some 108 pennate like Pseudonitzchia sp. The phytoplankton richness and diversity in this season was that 109 corresponding to natural ecosystems (Shannon index of 2.5).
In autumn campaigns (October) only two taxon were the most abundant in plankton 111 composition, the chlorococcal species of Chlorella (78%) and Scenedesmus (17%). Dinoflagellates, 112 diatoms and Cryptophyceae taxons represented less than 5% with a Shanon index of 1.

114
Conductivity was relatively high and was attributed to high evapotranspiration rates (4.3 mm) 115 and proximity to the sea (sea breeze). Decreases in conductivity occurred after precipitation events.

116
Dissolved oxygen was a critical parameter of water quality, with concentrations below 4.3 mg/L 40% 117 of the times studied. The effect of oxygen depletion was observed on fishes living in the lake,

118
Gambusia sp., which were observed staying at the most shallower depth. However, no mortality was 119 observed. Ammonia levels reached toxic concentrations for fishes three times, but again no mortality 120 was observed. Previously reported fish mortalities affected big fishes such as Mugil cephalus.

156
In this study we have focused on water quality after the first phase of La Goleta lake restoration 157 measures. We consider that the most appropriate parameters that shoud be mid-term monitored are 158 dissolved oxygen, ammonia and total chlorophyll a. These variables are easy to measure and the 159 information that they give us for water lake management very important. We also think that 160 monitoring phytoplankton blooms is clue as fish mortalities that could not be explained by oxygen 161 Journal Name 2016, x, x 5 of 5 depletion or ammonia toxicity could be due to phytoplankton toxins.
[12] monitored cyanobacteria 162 blooms and the associated cyanotoxin production for 14 months on a monthly basis in an urban lake.

163
However, we believe that this frequency may be not enough due to the high temporal variability that 164 characterizes phytoplankton. So, we recommend to the management authorities (the Council in this 165 case) to analyse phytoplankton when a bloom is detected by high chlorophyll a levels.
[1] suggested 166 to monitor chl a and TP concentrations as the most critical water quality variables for eutrophic lakes.

167
We recommend for urban lake monitoring the diagnosis of the main limitant nutrient, because this 168 can strongly influence the phytoplankton composition and abundance. At present, La Goleta lake 169 shows a higher potential limitation of nitrate rather than phosphate limitation.

170
Runoff from the watershed exerts significant influence on urban lakes and thus inflow nutrient 171 reduction is critical for the control of eutrophication [1]. Thus, the second phase of La Goleta lake 172 restoration measures will target pollutants and organic matter inflow. Sediments are an internal 173 nutrient source that will be also targeted, the mitigation with macrophytes such as vetiver is planned.

174
After second phase of restoration measures sediments quality will also be monitored. Degraded

198
The following nutrients were analysed in all the samples: nitrate, ammonium, dissolved 199 inorganic phosphorus (DIP), dissolved silicate (DSi) and total phosphorus (TP). Nutrients were 200 analysed colorimetrically using the methods of [17]. Nitrate was analysed using the Brucine Method 201 [18 ]. Dissolved inorganic nitrogen (DIN) was calculated as the sum of nitrate, nitrite and ammonium.

202
Chemical oxygen demand (COD) and Biochemical oxygen demand (BOD5) were analysed once to 203 estimate organic pollution of water.

204
In order to analyse the phytoplankton communities both epifluorescence and Uthermhol 205 microscopic counting methods were used. Epifluorescence was used to identify the pico and 206 nanoplanktonic cells size [19]. Samples contained in 250 mL glass bottles were fixed with 207 glutaraldehyde until reaching a final concentration of 2% [20]. Samples were filtered with Millipore 208 GTTP membranes (pore size 0.2 µm). Finally, a cover glass was placed on top of the filter [21]. The