Anthropogenic nitrogen impacts in littoral ecosystems: Quantification with N stable isotopes using long-living macroalgae

UNIVERSIDADE DE SANTIAGO DE COMPOSTELA FACULTAD DE BIOLOGÍA anthropogenic nitrogen impacts in littoral ecosystems: quantification with n stable isotopes using long-living macroalgae Memoria que presenta Inés González Viana para optar al grado de Doctora en Biología Fdo. Inés González Viana Santiago de Compostela, Noviembre 2014 Dr. Antonio Bode Riestra, Investigador en el Centro Oceanográfico de A Coruña del Instituto Español de Oceanografía, CERTIFICA Que la presente memoria titulada “Anthropogenic nitrogen impacts in littoral ecosystems: Quantification with N stable isotopes using long-living macroalgae”, presentada por Dña. Inés González Viana para optar al grado de Doctora en Biología, ha sido realizada bajo mi dirección, cumpliendo las condiciones exigidas para su presentación, la cual autorizo. Para que así conste a los efectos oportunos firmo la presente en Santiago de Compostela, a 7 de noviembre de 2014. Fdo. Dr. Antonio Bode Riestra. Director de la Tesis Doctoral. Dr. Alejo Carballeira Ocaña, Catedrático Ecología del Departamento de Biología Celular y Ecología de la Facultad de Biología de la Universidad de Santiago de Compostela y tutor de la presente memoria de Tesis Doctoral titulada: “Anthropogenic nitrogen impacts in littoral ecosystems: Quantification with N stable isotopes using long-living macroalgae”, de la que es autora Dña. Inés González Viana y que ha sido dirigida por el Dr. Antonio Bode Riestra, investigador del Centro Oceanográfico de A Coruña del Instituto Español de Oceanografía AUTORIZA su presentación en la Universidad de Santiago de Compostela para su lectura y defensa. Para que así conste, firmo la presente en Santiago de Compostela a 7 de noviembre de 2014 Fdo. Dr. Alejo Carballeira Ocaña Tutor de la Tesis Doctoral. This research was funded by projects ANILE (CTM2009-08396 and CTM201008804-E) of the Plan Nacional de I+D+i (Spain), and RADIALES of the Instituto Español de Oceanografía (IEO, Spain) To carry out this Doctoral Thesis, Inés González Viana was supported by a predoctoral fellowship from the Ministerio de Economía y Competitividad (Spain), 2010-2014 During the development of this Doctoral Thesis, a research stay was done at The Ecosystems Center, Marine Biological Laboratory, Woods Hole, MA, USA, funded by the Ministerio de Economía y Competitividad (Spain) Chapter 1. General Introduction Chapter 2. Stable nitrogen isotopes in coastal macroalgae: Geographic and anthropogenic variability Chapter 3. Ecology of Fucus vesiculosus (Phaeophyceae) at its southern limit of distribution: Growth and production of the early stages of development Chapter 4. Growth and production of new recruits and adult individuals of Ascophyllum nodosum in a non-harvested population at its southern limit (Galicia, NW Spain) Chapter 5. Experimental assessment of the macroalgae Ascophyllum nodosum and Fucus vesiculosus for monitoring N sources at different time-scales using stable isotope composition Chapter 6. Variability in δ15N of intertidal brown algae along a salinity gradient: differential impact of nitrogen sources Chapter 7. General Discussion Chapter 8. General Conclusions References Summary (Resumen) Acknowledgements (Agradecimientos) Contents 11 27 49 73 95 119 143 153 157 185 201 Tracing anthropogenic inputs Anthropogenic loads often increase the δ15N of dissolved inorganic nitrogen (DIN) of a system because they encourage denitrification (Mariotti et al. 1981, 1984; Sigman et al. 2009). The main sources of anthropogenic inputs are nitrogenous fertilizers, and animal or sewage wastes (Heaton 1986). DIN derived from sewage and farm wastes may be significantly enriched in 15N as a result of volatilization and microbial processing of the nitrogen in solution (Sweeney and Kaplan 1980, Rau et al. 1981, Heaton 1986, Van Dover et al. 1992). Animal waste has a high ammonium-N component; the volatilization of ammonia gas causes the loss of 14N with a resulting enrichment of 15N in the remaining nitrogen, which is converted to nitrate. In contrast to these anthropogenic sources, agricultural activity generally results in low δ15N values in nitrate of the adjacent surface waters. Fertilizers have a wide range of values because of the variety of source materials used to create them. Ammonium-based fertilizers can be converted to nitrate by the bacterially mediated process of nitrification (Kendall and Caldwell 1998). The volatile loss of ammonia from organic and inorganic fertilizers, and denitrification in soils leads to higher δ15N-NO3 - values of nitrate in rivers receiving large amounts of N from agriculture, compared with natural watersheds (Flipse and Bonner 1985, Mayer et al. 2002). In contrast, ammonium fertilizers, which are produced by fixation of atmospheric N (by the Haber-Bosch process), show small differences in δ15N as a result of small fractionation during subsequent processing of the fixed N (Flipse and Bonner 1985). All these pollutant sources in the environment have characteristic isotopic composition ranges (Fig. 1.2). These isotopic signatures can also vary while nutrients move from terrestrial to aquatic systems and along watersheds, and the N cycle may increase the δ15N values through subsequent nitrification, denitrification or volatilization processes (Bedard-Haughn et al.2003). Therefore, δ15N-DIN can be used to interpret the source of environmental contamination (Heaton 1986, Lindau et al. 1989, Aravena et al. 1993). Macroalgae as indicators of N sources: N stable isotopes biomonitoring Even though δ15N-DIN values have been widely used for assessing anthropogenic N loadings (Kendall 1998, Mayer et al. 2002, Deutsch et al. 2005), in the last years several limitations of this method for monitoring purposes have arisen. While the determination 16 of isotopic values of dissolved nutrients help to identify nutrient sources, there exists high variability due to tidal exchanges and transient influx of nutrient concentrations in some coastal areas (Valiela et al. 1997). So given that nutrients may vary as a result of a wide range of factors, it is difficult to clearly identify all sources from instantaneous water measurements of isotopic values of DIN components. Moreover, stable isotope measurements on dissolved nutrients may not reflect what is actually available to primary producers, as saturation uptake kinetics or differential growth periods along the year have been observed (Lyngby 1990). Partly as a result of these limitations or the possibly higher effort of δ15N-DIN analysis compared to determinations in biota, increasing attention has been given to macroalgae as biomonitors of changes in the nutrient loads of coastal waters. Nearly all transformations of the N cycle are undertaken by marine organisms as part of their metabolism, either to obtain nitrogen to synthesize structural components, or to gain energy for growth (Gruber 2008). This internal cycling may result in different nitrogen reservoirs of intermediary biota with also characteristic isotopic signatures (Owens 1987). From an isotopic standpoint, all of the different sources of N discussed in the previous section may also alter the baseline δ15N of Ammonium fertilizer Nitrate fertilizer Urea Ammonium precipitation Nitrate precipitation Soil N Manure Sewage Uncontaminated ground water Ground water with nitrate sources Uncontaminated surface water Surface water with nitrate sources -20 -10 0 10 20 30 40 δ 15 N (‰) Figure 1.2. Box plots of δ15N values of NO3 - from various sources and sinks. Box plots illustrate the 25th, 50th and 75th percentiles; the whiskers indicate the 10th and 90th percentiles; and the circles represent outliers (adapted from Xue et al. 2009). Chapter 1: General introduction 17 the marine biota. Theoretically, direct detection of wastewater N in biota should provide a means to identify contributions of N to estuarine food webs before increased N availability leads to visible changes of population and community (McClelland and Valiela 1998b). Several studies have exploited the characteristic isotopic signatures of anthropogenic N sources in macroalgae (McClelland and Valiela 1998b, Costanzo et al. 2001, Gartner et al. 2002, Alquezar et al. 2013). The advantages of these primary producers are that they are easy to identify, sample and process, and they are widely distributed. They accumulate dissolved substances in proportion to ambient bioavailability, offering time-integrated measures of the exposure of the biomonitor to sources over a previous time period (Rainbow and Phillips 1993). As they are sessile, they can integrate water variations of a particular site. Attending to nutrition strategies, different groups can be distinguished, as ephemeral and perennial species (Martínez et al. 2012). Among ephemeral or fast-growing species, several green and red macroalgae can be included, as Ulva sp. All of them have in common their high nitrogen requirements (Pedersen and Borum 1997). This high nitrogen demands are associated with their high growth rates and, to a lesser extent, to their high tissue-N content (Hanisak 1983). Their capacity for N storage is also more limited than in perennial species, resulting in short-term storage that is gained through surge uptake. In contrast, perennial species, as fucoid macroalgae, are long-living species, and due to their slower growth rates, their N demand is also lower. This group of species has, in addition, higher capacity for utilizing internal N stores to cover their requirements for growth at low external N availability than the ephemeral macroalgae (Pedersen and Borum 1996). These differences in physiology and ecology have been shown to make a difference between N metabolism and requirements of both groups (Lobban and Harrison 1994). Extensively, these differences might influence the bioavailable N isotopic fraction reflected by the different species. Therefore, assessment of the relation of different species and stable isotopes needs to be separately tested. Fast-growing macroalgae have received large attention in order to find the linkages between anthropogenic sources of N and its influence on enhancing primary production (macroalgal blooms). Therefore a number of studies in the last years have focused on understanding how these opportunistic macroalgae and stable 1818 isotopes variability are related (Naldi and Wheeler 2002, Rogers 2003, Barile 2004, Cohen and Fong 2004, Lapointe et al. 2005, Dailer et al. 2010, Barr et al. 2013). Due to their seasonal living-strategy, these macroalgae may not be found year-round at some locations, so they might not be useful when studying the overall status of a particular site. In contrast, long-living macroalgae may not reflect sporadic nutrient changes, but may be suitable for long-term studies. The suitability of fucoid macroalgae for monitoring N loading have been also tested in several locations and under different N sources (Table 1.1). Despite the number of studies using δ15N values in brown macroalgae for monitoring purposes, there are almost no studies examining sources of variability affecting the isotopic values and net fractionation, especially in fucoid species (Umezawa et al. 2007, García-Sanz 2009). Therefore, the use of stable isotopes on these macroalgae to detect N loading has required some assumptions related to their ecology, physiology and fractionation processes. As shown by studies in other brown algae, species-specific chemical and physiological analysis, and understanding of their individual responses to various environmental factors are needed to interpret algal isotopic signatures (Umezawa et al. 2007). Among fucoid macroalgae, Fucus spp. and Ascophyllum nodosum are widely distributed at both sides of the Atlantic Ocean, from the White Sea to the Iberian Peninsula at the East coast, and from around 77° N to New Jersey on the western side (Baardseth 1970). They show apical growth, so the growing tip is the part of the frond that has been traditionally considered in monitoring studies (Table 1.1). In these studies the current trophic status of different environments were characterized by the macroalgal δ15N. Other interesting feature of their apical growth is the use of these species in retrospective studies. The δ15N values of different segments along the frond can be related with past growing periods, and hence with ambient N status during those periods, if their growth rate is known. Savage and Elmgren (2004) estimated the year of growth of different sections of the thallus of Fucus vesiculosus from the number of bifurcations. However, specific studies showed large local variability and thus the inconsistent relationship between the number of bifurcations and growth in this species (Knight and Parke 1950). In contrast, A. nodosum individuals produce annual gas bladders that have been used in other studies to delimitate different exposure periods to different contaminants (Heldal and Sjøtun 2010); Chapter 1: General introduction 19 Species δ 15 Nδ 15 N source Nsource Location Sampling period Part of the tissue n Reference Fucus spp. 1.27-10.41 -fish farms (44-2,250 t year -1 )Galicia, Spain July 2008tips (3 cm)~30Carballeira et al. 2013 Fucus spp. 5.6-5.9 - reference sites Galicia, Spain July 2008tips (3 cm)~30 F. vesiculosus 7.18-10.47 -fish farm (2, 250 t year -1 )Galicia, Spain July 08, 09, 10 retrospective study 3Carballeira et al. 2014 F. vesiculosus 4.75-6.77 -reference site Galicia, Spain July 08, 09, 10 retrospective study 3 A. nodosum 9.4 9.3 WWTP outfalls, septic systems Oyster River, Great Bay, NH, USA - - - Cole et al. 2004 F. vesiculosus 5.6-7.9 3.7-9.0 63% agriculture, 24%forests, 7%urban area Warnow River Estuary, Germany Feb-March 2004 tips (1-1.5 mm)5 Deutsch and Voss 2006 F. vesiculosus 7.3 6.7-9.7 63% agriculture, 24%forests, 7%urban area Warnow River Estuary, Germany May 2004 tips (1-1.5 mm)5 F. vesiculosus 11-13 - < 15 km from treated sewage of 250,000 inhab Himmerfjärden Bay, Sweden May tips - Hobbie et al. 1990 F. vesiculosus 4.6 - at 35 km from treated sewage of 250,000 inhab Himmerfjärden Bay, Sweden May tips - F. spiralis 9.2 - STP Hope Island, Narragansett Bay, RI, USA September 2006 - 1 Oczkowski et al. 2008 F. spiralis 8.7 - STP Dutch Island, Narragansett Bay, RI, USA September 2006 - 1 Fucus sp. 7.7 - STP Beavertail, Narragansett Bay, RI, USA September 2006 - 2 A. nodosum 9.8 - STP Whale Rock, Narragansett Bay, RI, USA September 2006 - 1 F. spiralis 8.4STP Whale Rock, Narragansett Bay, RI, USA September 2006 - 2 F. spiralis 9.6 - STP Gould Island, Narragansett Bay, RI, USA September 2006 - 1 A. nodosum 8.1 - STP Castle Hill, Narragansett Bay, RI, USA September 2006 - 1 F. distichus 7.9 - STP Castle Hill, Narragansett Bay, RI, USA September 2006 - 1 Fucus sp. 8.2 -offshore influence Block Island Great Salt Pong, RI, USA September 2006 - 1 F. vesiculosus 9.5±1culture, pasture, STP Marebes-Oléron Bay, France Jan-May 2006tips (2 cm) 3 Raimonet et al. 2013 F. vesiculosus 9.8±1.3 8.5±3.9culture, pasture, STP Marebes-Oléron Bay, France Jan-May 2006tips (2 cm) 3 F. vesiculosus 6.4±2.37.4±0.8reference oceanic influence Marebes-Oléron Bay, France Jan-May 2006tips (2 cm) 3 F. serratus 10.5±0.3culture, pasture, STP Marebes-Oléron Bay, France Jan-May 2006tips (2 cm) 3 F. serratus 8±2.1 culture, pasture, STP Marebes-Oléron Bay, France Jan-May 2006tips (2 cm) 3 F. serratus 2.1±3.37.4±0.8reference oceanic influence Marebes-Oléron Bay, France Jan-May 2006tips (2 cm) 3 F. vesiculosus 25.7±3.8 - eutrophic tidal estuary The Westerschelde Estuary, The Netherlands Jan-Feb 1998 - 2Riera et al. 2000 F. vesiculosus 6.3±0.2 -tidal bay The Oosterchelde Estuary, The Netherlands Jan-Feb 1998 - 2 F. vesiculosus 8-9 38 <4 km from STP Himmerfjärden Bay, Sweden May 1999 tips (~2 cm) 3 Savage and Elmgren 2004 F. vesiculosus 5-6 38 >4 km from STP Himmerfjärden Bay, Sweden May 1999 tips (~2 cm) 3 F. vesiculosus 3-4 4 14 km (coastal reference station) Himmerfjärden Bay, Sweden May 1999 tips (~2 cm) 3 F. vesiculosus 10.5-14 24-38 ~1km Himmerfjärden Bay, Sweden May 2002 retrospective study 3 F. vesiculosus 3.4-3.84 coastal reference station Himmerfjärden Bay, Sweden May 2002 retrospective study 3 F. ceranoides 2.17-14.57 -sewage, upwelling, agriculture Galicia, Spain July 1990-2007 tips (3 cm)~30Viana et al. 2011 F. spiralis 6.12-11.33 - sewage, upwelling, agriculture Galicia, Spain July 1990-2007 tips (3 cm)~30 F. vesiculosus 3.96-11.23 - sewage, upwelling, agriculture Galicia, Spain July 1990-2007 tips (3 cm) ~ 30 Table 1.1. Overview of studies using δ15N values in Fucus spp. and A. nodosum to track anthropogenic inputs in coastal systems. 20 although there is also some uncertainty in determining when the first gas bladder appears (Cousens 1984). Thus, the growth, number of bifurcations and the timing in the appearance of the gas bladders need to be quantitatively tested for a feasible use of Fucus spp. and A. nodosum fronds in punctual and retrospective studies. The lack of site-specific growth studies, determines the necessity of quantitative growth measurements in the areas of interest, especially due to the large type of habitats where these species live. Apart from the species-specific growth rates, it is necessary to understand the N metabolism along the macroalgal thallus. The main assumption of retrospective studies is that only the growing tips of the thallus take up nitrogen and, therefore, the isotopic composition of a given section of the thallus would reflect the isotopic composition of the dissolved nitrogen in the surrounding water at the time of growth (Savage and Elmgren 2004). Therefore, some questions need to be tested to fully interpret the data obtained in these studies. First, Fucaceae do not have a specific transport tissue, but the pores of the sieve plates would enable a continuous system of cytoplasm for the translocation of materials longitudinally (Moss 1983). If nitrogen transport along the thallus exists, it would directly affect the retrospective identification of past nitrogen sources. Second, most studies assume that the isotopic composition of tissues does not change for at least several months, given that these species generally show low variability in δ15N values at monthly time scales (Gartner et al. 2002) but no data of N-specific uptake and turnover rates are available for most Fucacean species. As they are long-living perennial species, it is also likely that isotopic signatures in macroalgae may be confounded by a variety of other factors in the environment, as seasonal effects or responses to local factors. For instance, the coast of the southern distribution limit of these species is eutrophized by seasonal upwelling processes (Álvarez-Salgado et al. 1997), while the Baltic Sea is known to be mainly eutrophized by the increasing anthropogenic influence (Bonsdorff et al. 1997). If the sources of isotopic variability at large environmental or geographical scales are not known, it would be difficult to exploit the isotopic values of these species and relate it to anthropogenic pressures in regional-scale studies. Chapter 1: General introduction 21 They also live in the intertidal area of a wide variety of environments; therefore they are subject to a wide spatial and seasonal variability of abiotic factors and sources (e.g. salinity, temperature). This variability might significantly influence isotopic values in their tissues, but whether the variability observed in some studies is due to the sources or to the N metabolism in macroalgae (i.e. net fractionation of all processes within macroalgal tissues) is not known. Simultaneous measures of both macroalgae and seawater need to be done to detect the reasons of this variability (Deutsch and Voss 2006). HYPOTHESIS AND GENERAL OBJECTIVES Taking into account the present knowledge on the aplication of δ15N in Fucaceae for monitoring anthropogenic N inputs to the coastal zone, and the information that is lacking, the following hypothesis was established: • The proportion of N stable isotopes in the structural tissues of long-living macroalgae reflects the different utilization of natural and anthropogenic sources of nitrogen at long-time scales (months, years). To validate the former hypothesis, the following principal and partial objectives were established: • To determine the relation between the natural abundance of N stable isotopes in macroalgae and the origin of the sources of available N. 1. To determine the geographical variability of the isotopic composition of the macroalgae Ascophyllum nodosum and Fucus spp. in relation to natural (e.g. upwelling) and anthropogenic (e.g. urbanization) factors. 2. To determine the growth rates of A. nodosum and Fucus vesiculosus for their possible application to retrospective studies of nitrogen inputs. 3. To determine the permanence period of N in different parts of the macroalgal frond. 4. To quantify the anthropogenic N impact on intertidal macroalgae. 22 GENERAL STRUCTURE OF THE PhD THESIS The data of the present dissertation were gathered during the years 2010-2013 within the project “Anthropogenic nitrogen inputs to littoral ecosystems: basis for monitoring using stable isotopes” (ANILE; CTM2009-08396 and CTM2010.08804-E). The general aim of this project was a precise quantification of natural and anthropogenic nitrogen inputs to littoral ecosystems. The specific objectives were to obtain information on the variability in the concentrations of dissolved nitrogen forms, and determine the isotopic signatures in the main components of plankton and benthos compartments. The objectives of the present thesis were within the partial objectives of this project. To fulfill the previously numbered objectives, the analysis of these data was divided in different chapters that are structured as follows: In Chapter 2 the natural and anthropogenic variability of N stable isotopes in macroalgae ( F. vesiculosus and A. nodosum ) is analyzed along a biogeographic gradient in an upwelling area (NW Spain). The influence of the upwelling seasonal process along the coast is based on previous studies, which showed that the influence is weaker from south to north. The anthropogenic influence is estimated based on the population size of the urban nuclei. The third and fourth chapters aim to study the growth rates of both target species ( F. vesiculosus and A. nodosum ) with the aim of using the resulting growth curves in retrospective studies. The appearance of gas bladders in A. nodosum is also studied together with other aspects of the ecology of both species. In the fifth chapter different assumptions when using macroalgae ( F. vesiculosus and A. nodosum ) as retrospective biomonitors are tested. First, the study focused on elucidating the turnover times of different parts of the tissue under different experimental and natural sources of water. Second, the existence of uptake capacity of growing and non-growing parts of the frond was determined. In addition, the possible existence of long-distance transport of nitrate along the thallus was tested in both species. The understanding of processes affecting the variability of δ15N in the selected macroalgae gathered in previous chapters were applied in the sixth chapter. A Chapter 1: General introduction 23 local study of the influence of different terrestrial (i.e. river discharge) and natural (i.e. upwelling) N sources to these macroalgae was done at a local scale. In this chapter, various spatial and temporal scales of variability are considered. The seventh chapter presents a synthesis of the main results and discussion within this dissertation. A general discussion of the implications of our findings and some future prospects are also suggested. Finally, the last chapter gathers the main conclusions of this PhD Thesis. 24 Material and methods Sampling Samples were collected in the intertidal along the coast of NW Spain at sites representative of environments with variable influence of the upwelling and in a large range of urban influence (Fig. 2.1). As upwelling in the northern coast is generally weaker than in the western coast (Botas et al. 1990), an arbitrary reference point located at the sea discharge point of the River Miño (Fig. 2.1) was used to compute the distance along the coast between each sampling site and this reference point. This distance was intended to indicate the lower input of new nitrogen by the upwelling in the northern coast (Mar Cantábrico, zone I in Fig. 2.1) compared to those in the western coast (Galicia). In the latter, two zones were considered to investigate potential differences between Rías Baixas (zone III) and other rias (zone II). Sampling sites covered a large range of urban population influence in the watershed (from ~240 to ~246,000 inhabitants) according to Spanish Official Population Census (http://www. ine.es/inebase). Sampling surveys were carried out mostly during spring and summer 2010 and 2011, but some samples from 2006 were added to complete the range of geographic or urban population values (Table 2.1). Two species of Phaeophyceae (brown algae) were selected: Ascophyllum nodosum and Fucus vesiculosus . The species were present at 12 and at 26 sites respectively, and they were cohabiting at 11 sites. Three individuals of each macrophyte species fixed to the substrate were collected from the meso-littoral zone when emerged. Apical parts of the specimens (1 cm) were used for analysis of the stable nitrogen composition. Samples were rinsed with Milli Q water to remove sediments and other material and frozen (-20 °C) before processing. Samples were defrosted and dried (50 °C) until constant weight, before grinding into a homogeneous powder. Samples of surface water were collected concurrently with macroalgae. Salinity was measured in situ with a portable conductivity meter (YSI Model 30). Water samples were poisoned with HgCl2 (0.05% final concentration) to prevent microbial alteration and stored in tightly caped Pyrex flasks. 32 Table 2.1. Mean (±se) values of total nitrate (NO3 -+NO2 -) and ammonium (NH4 +) concentrations and δ15N in water and macrophyte samples at the sampling sites. Salinity (S) and the number of inhabitants in the watershed (population) are also indicated. Code is the number of each site in Fig. 2.1. Concentration (µM) 15N Code Site Latitude Longitude Date Population S NO3- + NO2NH4+ NO3- + NO2NH4+ A. nodosum F. vesiculosus 1 El Sardinero 43.48145 -3.78715 11/05/2011 141,269 34.3 2.15±0.02 - - - - 5.8±0.3 2 Toró 43.41743 -4.74270 12/05/2011 276 34.9 3.63±0.51 - - - - 4.5±0.1 3 El Sablón 43.42247 -4.75226 12/05/2011 5,358 33.8 3.63±0.04 - - - - 5.5±0.1 4 La Griega 43.50288 -5.26320 06/08/2010 3,878 33.9 4.33±0.60 ≥10 5.1±0.3 - - 4.5±0.2 5 El Puntal 43.52605 -5.38812 06/08/2010 239 32.7 1.80±0.25 ≥10 4.0±0.2 - 6.1±0.3 - 6 Xivares 43.56827 -5.71207 05/08/2010 2,675 34.2 5.36±0.75 ≥10 4.8±0.3 - - - 7 S. Juan de la Arena 43.55705 -6.07709 16/04/2010 1,970 5.3 19.68±2.74 4.03±0.55 4.5±0.2 - - 10.5±0.1 8 Navia 43.55214 -6.72481 16/04/2010 8,906 1.8 22.37±3.12 2.28±0.31 3.7±0.2 - - 6.2±0.2 9 Figueras 43.53794 -7.02360 16/04/2010 3,845 29.8 14.80±2.06 2.82±0.38 18.0±1.0 - 5.1±0.2 6.9±0.1 10 Ribadeo 43.53539 -7.03596 31/07/2010 9,983 29.3 9.50±1.32 ≥10 6.2±0.3 - - 6.8±0.6 11 Foz 43.56468 -7.24599 31/07/2010 13,214 27.2 14.40±2.01 2.75±0.37 3.7±0.2 - 5.7±0.3 6.4±0.1 12 Cedeira 43.66007 -8.05606 16/08/2010 7,465 31.7 4.28±0.60 ≥10 6.5±0.3 - - 13.8±0.6 13 Vilarrube 43.64518 -8.08386 16/08/2010 363 28.6 5.83±0.81 3.67±0.50 2.5±0.1 - 7.8±0.2 8.2±0.0 14 A Graña 43.47893 -8.26019 25/07/2010 74,273 34.5 4.14±0.58 ≥10 2.9±0.2 0.3±0.1 7.7±0.1 7.2±0.4 15 Cabanas 43.41146 -8.17255 17/08/2010 11,793 29.6 7.20±1.00 ≥10 4.9±0.3 - 6.2±0.3 6.9±0.2 16 Mera 43.38247 -8.34397 18/04/2011 32,947 32.8 14.74±2.05 3.51±0.48 4.7±0.3 - - 8.2±0.0 17 O Burgo 43.32770 -8.37034 18/04/2011 83,691 26.5 39.38±5.49 4.72±0.64 3.3±0.2 - 8.9±0.1 9.5±0.0 18 A Coruña 43.36916 -8.38836 16/02/2006 243,349 - 1.92±0.27 4.85±0.66 - - - 8.0±0.2 19 Bens 43.36926 -8.45777 26/07/2010 246,056 35.6 5.93±0.83 ≥10 4.8±0.3 0.8±0.1 - 4.8±0.2 20 Caión 43.31825 -8.60719 15/02/2006 661 - 3.10±0.43 7.47±1.02 - - - 4.7±0.0 21 Pontevedra 42.42799 -8.65340 24/07/2010 81,756 26.4 11.80±1.64 ≥10 4.0±0.2 -1.6±0.1 - 8.9±0.2 22 Placeres 42.40659 -8.68541 10/08/2010 16,996 35.3 3.29±0.46 4.25±0.58 4.7±0.3 - 7.2±0.2 6.5±0.2 23 Aguete 42.37571 -8.72958 10/08/2010 1,075 35.6 1.40±0.20 ≥10 4.8±0.3 - - 6.4±0.1 24 Soutomaior 42.34022 -8.61412 24/07/2010 6,867 26.5 5.22±0.73 ≥10 5.7±0.3 2.4±0.1 -1.6± 0.4 1.6±1.3 25 Cesantes 42.29945 -8.61677 24/07/2010 30,001 35.5 5.61±0.78 ≥10 6.5±0.3 2.6±0.2 9.9±0.3 8.7±0.2 26 O Latón 42.27885 -8.70626 28/08/2010 19,014 2.4 7.10±0.99 ≥10 19.6±1.0 - - 9.5±0.5 27 Meira 42.27654 -8.71091 18/02/2006 18,415 - 2.27±0.32 4.12±0.56 - - 10.1±0.1 8.0±0.2 28 Ramallosa 42.12180 -8.81998 24/07/2010 18,021 32.5 10.85±1.51 ≥10 5.3±0.3 1.9±0.1 10.1±0.9 10.2±0.1 33 Chapter 2: Geographic and anthropogenic variability Chemical analysis Nitrate, nitrite and ammonium were determined in the laboratory using segmented flow analysis (Braun-Luebbe AAII) following the procedures of Grasshoff et al. (1983). Sensitivity was 0.05, 0.01 and 0.04 µM for nitrate, nitrite and ammonium, respectively. Precision (se of 3 replicates) was better than 14% of the mean value for any of the nitrogen species. Ammonium values >10 µM were excluded from further analysis because of suspect contamination of samples during processing, as values reported for coastal waters in the study region do not exceed 10 µM (e.g. Bode et al. 2011b). The isotopic composition of total nitrate (NO3 -+NO2 -) was determined by previous conversion into ammonium and later recovery of ammonium on a solid phase. The procedure is an adaptation of the diffusion method (Sigman et al. 1997) involving the incubation of samples in two steps. In this case the resulting ammonium was collected on a small disk of glass-fiber filter placed in the gas headspace of the diffusion flask (Slawyk and Raimbault 1995). First, aliquots of the samples were incubated (50 °C, 1 week) in the same collecting flask without cap to reduce the volume and concentrate nitrate. Ashed MgO was added to raise pH above 9.7 to remove ammonia by volatilization. In the second step (50 °C, 2 weeks), ashed Devarda´s alloy was added to the reduced volume sample to convert nitrate and nitrite into ammonium. The high pH (>11) of the mixture ensured also the conversion of ammonium into ammonia gas that was collected on a sterilized glass-fiber disk (Whatman GF/F), acidified with 0.5 ml of 0.25N H2SO4 and hooked on a needle fixed to the inner side of the flask cap. Care was taken to ensure that the filter disk did not contact the liquid sample. This extraction procedure does not allow separation between NO3 - and NO2 - therefore the values reported are the combined isotopic signatures of total nitrate (Ahad et al. 2006). After the second incubation step the disk filters were dried and prepared for isotopic analysis. The stable isotope composition of ammonium was determined in another aliquot of the water samples by an adaptation of the diffusion method (Holmes et al. 1998). This method involves gas-phase diffusion as described for the second step of the total nitrate extraction. In all cases corrections for isotopic fractionation during the whole incubation and diffusion steps were made (Holmes et al. 1998). The measured values of natural abundance of dissolved inorganic nitrogen were retained for further analysis when the ammonium recovery after the diffusion procedure exceeded 45% and isotopic fractionation of internal standards was within 1‰ of values estimated from the empirical equation in Holmes et al. (1998). 34 Stable isotopes The natural abundance of stable nitrogen isotopes was determined in macroalgae and water samples (total nitrate and ammonium). For macroalgae, 2.5 mg of dry sample was analyzed to ensure a minimum of 10 µg of N. For water samples, 1 ml of 4 mM-N (NH4)2SO4 was added to each sample during the diffusion phase to ensure the detection limit was achieved. Samples were placed in tin capsules and introduced into an isotope-ratio mass spectrometer (Thermo Finnigan Mat Delta Plus) via an element analyzer (Carlo Erba CHNSO 1108). Isotopic results are expressed in delta notation: δ15N = [(15Nsample:14Nsample/15Nstd:14Nstd)-1] x 1000 where the standard (std) for δ15N is atmospheric N2. Precision (se of 5 replicates) was better than 0.05‰ for either IAEA-N-2, IAEA-N-1 or IAEA-NO-3 standards. The coefficient of variation of triplicate sample aliquots was always <2%. Statistical procedures Relationships between variables were first analyzed using non parametric correlation (Spearman ρ). Further analyses were made using linear regression after excluding outliers exceeding 1.5 times the interquartile range. In the case of salinity vs. dissolved nitrogen concentrations and macroalgal δ15N vs. geographical distance, product-moment regression was used because either the error in estimating the salinity was much lower than the error for dissolved nitrogen or because the resulting slope was further employed to account for systematic variability in δ15N with geographical distance (Sokal and Rohlf 1981). In the case of the comparison of δ15N between the two macroalgal species standard major axis was used because both variables were measured with the same type of error (Sokal and Rohlf 1981). In this latter case, the obtained regression parameters were compared with the line of slope 1 and zero intercept by a t -test (Warton and Ormerod 2007). The relative contribution of geographical distance and population size to δ15N was estimated as the sums of squares (Type I) obtained with an ANOVA design including two population size classes (larger and smaller than 15x103 inhabitants, respectively) with distance as covariable. Differences between sampling zones or classes of population size were further analyzed by non parametric Kruskal-Wallis (K-W) test (Sokal and Rohlf 1981). 353535 Chapter 2: Geographic and anthropogenic variability Results Dissolved inorganic nitrogen Total nitrate concentration in the samples ranged from 1.40 to 39.38 µM, while ammonium (excluding >10 µM values) ranged from 2.28 to 7.47 µM (Table 2.1). Total nitrate was negatively correlated with salinity in most samples (Spearman ρ =-0.682, P<0.001, n=24) except at O Burgo, where nitrate reached ca. 40 µM (Fig. 2.2). In contrast, ammonium was not correlated with salinity (P>0.05). These relationships with salinity suggest large potential contributions of nitrate from freshwater in most of the studied area but variable inputs of ammonium unrelated to freshwater discharges. 30 40 50 6 8 µ M) NH4 NO3+NO2NO 3- +NO 2NH 4+ 10 20 30 2 4 6 NO 3- + NO 2- (µM) NH 4+ (µM) NO3+NO2 NO 3 +NO 2 0 10 0 2 0 10 20 30 40 NO Salinity Figure 2.2. Linear relationships between ammonium (NH4 +, black squares) or total nitrate (NO3 -+NO2 -, gray circles) and salinity in water from the sampling sites. The point encircled was an outlier (>1.5 times the interquartile range) not used in the estimation of the regression line (Spearman ρ=-0.682, P<0.001). 36 Because of rapid contamination with ambient ammonia during the analytical preparation steps stable isotope composition of dissolved nitrogen was determined with confidence in a subset of samples only (Table 2.1). Total nitrate δ15N varied between 2.5 and 19.6‰ while δ15N ammonium ranged from -1.6 to 2.6‰ (Table 2.1). When measured concurrently δ15N of ammonium and δ15N of total nitrate were correlated (Spearman ρ=0.943, P<0.01, n=6). The highest nitrate value corresponded to the sample from O Latón (Code 26), collected at the discharge outlet of a Water Treatment Plant, but a large value was also observed in Figueras (Code 9), in this case not obviously related to residual water discharges. Values of nitrate δ15N for marine waters (salinity >35) were near 5‰. δ15N in macroalgae Stable isotope composition of F. vesiculosus and A. nodosum was significantly correlated (Spearman ρ=0.806, P<0.01, n=10). The resulting regression line did not differ from a line with slope 1 and intercept 0 (P<0.05) indicating that the isotopic composition of these species was equivalent for a given site (Fig. 2.3). In contrast, macroalgal δ15N was not correlated with either dissolved inorganic nitrogen concentrations, salinity or isotopic composition (Fig. 2.4). Geographic variability in δ15N Macroalgal δ15N varied according to the geographical location of samples (Fig. 2.5). Both species showed a linear decrease in δ15N with the distance from the reference point in the River Miño (Fig. 2.5a). The slope of the regression lines indicated a change of δ15N of 0.3 and 0.4‰ per 100 km of coastline for F. vesiculosus and A. nodosum respectively (Table 2.2). In contrast a significant relationship was not found between dissolved nitrogen concentrations or δ15N of total nitrate with distance, as exemplified by total nitrate concentration (Fig. 2.5b). No significant differences resulted either when considering the sampling zones (I, II and III) in a K-W test (P>0.05). 8 12 u m 0 4 8 δ15NA. nodosum -4 0 -4 0 4 8 12 δ1 5 δ 15 N F vesiculosus -4 0 4 8 12 δ15NF. vesiculosus Figure 2.3. Relationship between stable isotope composition of Ascophyllum nodosum and Fucus vesiculosus sampled at the same locations. The regression line computed without the outlier (open circle, >1.5 times the interquartile range) is significant and with zero intercept (Spearman ρ=0.806, P<0.01) while the slope is non-significantly different from 1. 37 Chapter 2: Geographic and anthropogenic variability Samples of F. vesiculosus collected inside the rias and estuaries (as shown in Fig. 2.1) had higher δ15N values than samples collected in open coastal sites (K-W test, P<0.01). Mean (±se) values for rias and coastal sites, after correction for the geographic variability using the slope in Table 2.2, were 9.1±1.1‰ (n=17) and 7.6±1.1‰ (n=7), respectively. Figure 2.4. Biplots of macroalgal δ15N and concentrations of total nitrate (a) and ammonium (b) or δ15N in total nitrate (c) and ammonium (d). None of the relationships is significant (Spearman ρ, P>0.05). 38 Variability of δ15N with human population The geographic variability accounted for more than half of total variance in δ15N for both species (Fig. 2.6). However, the size of the human population in the watershed was also an important factor for δ15N, particularly for A. nodosum . The isotopic values of both macroalgae, after removal of the geographic trend using the equations in Table 2.2, increased non-linearly with the size of the human population in the watershed (Fig. 2.7). -4 0 4 8 12 16 0 10 20 30 40 50 15N in macroalgae NO3-+NO2-(µM) F. vesiculosus A. nodosum a -4 0 4 8 12 16 0 2 4 6 8 10 15N in macroalgae NH4+(µM) b -4 0 4 8 12 16 0 5 10 15 20 15N in macroalgae 15N(NO3-+NO2-) c -4 0 4 8 12 16 0 1 2 3 4 15N in macroalgae 15N(NH4+) d -4 0 4 8 12 16 0 400 800 1200 15N in macroalgae relative distance (km) A. nodosum F. vesiculosus a 0 10 20 30 40 50 0 400 800 1200 NO3-+ NO2-(µM) relative distance (km) b Figure 2.5. Variability of δ15N in macroalgae (a) or total nitrate (b) with the relative distance of sampling locations to the River Miño discharge point (see Fig. 2.1). The regression lines for Ascophyllum nodosum (Spearman ρ=-0.855, P<0.01) and Fucus vesiculosus (Spearman ρ=-0.590, P<0.01) are indicated. Outliers of δ15N (>1.5 times the interquartile range and not used in the estimation of regression lines) are enclosed in circles (a) while the corresponding inorganic nitrogen concentrations are shown as open dots (b). Variability in δ15N was largest at small population sizes (<50x103 inhabitants) with clear outliers with unusually large or small values. At the three sites influenced by large populations (>100x103 inhabitants) δ15N values in F. vesiculosus (as A. nodosum was not found at these sites) did not follow the increase observed at lower populations. In turn, the distribution of the human population has no relationship with the geographical gradient found for macroalgal δ15N (no significant correlation between population size and distance). In any case, and excluding the outliers, both species showed significantly higher δ15N values at population sizes larger than 15x103 inhabitants (Fig. 2.8, K-W test, P<0.05). 3939 Chapter 2: Geographic and anthropogenic variability Discussion Natural variability of nitrogen sources Differences in both concentration and δ15N values of nitrate were expected in the NW Spanish coast because of the varying influence of the upwelling, as nitrate from the Eastern North Atlantic Central waters is the main natural source of nitrogen for primary production in shelf waters of this area (Botas et al. 1990, Casas et al. 1997, ÁlvarezSalgado et al. 2002). Instead, our results indicated no significant spatial variability pattern of nitrate concentrations or δ15N. Nitrate was the main form of dissolved inorganic nitrogen and its highest concentrations were found in estuarine waters, suggesting a significant input from freshwater. However, given the low flow of rivers in 50 60 20 30 40 50 variance % F. vesiculosus A. nodosum 0 10 20 var A. nodosum 0 distance population class error Figure 2.6. Contribution of distance to the reference point (as covariable) and human population (as fixed factor with two levels: larger and smaller than 15x103 inhabitants, respectively) to the variance of δ15N in Fucus vesiculosus and Ascophyllum nodosum . The error term includes the remaining variability not accounted for by all other components. The outliers in Fig. 2.5 were not included in the analysis (ANOVA, P<0.05 for all components). 40 Table 2.2. Linear regression parameters (δ15N = a + b distance) of the variation of δ15N in Fucus vesiculosus and Ascophyllum nodosum with the distance in km to the River Miño. P: significance, n: number of data points, se: standard error. The outliers in Fig. 2.5 were excluded from the estimation. species a±se b±se r P n F. vesiculosus 8.774±0.530-0.003±0.001 0.639 0.001 23 A. nodosum 9.889±0.610 -0.004±0.001 0.8190.002 11 Figure 2.7. Variability of δ15N in Fucus vesiculosus (a) and Ascophyllum nodosum (b) with the size of the human population in the watershed. The curves are polynomial (a) or lineal (b) fits and 95% confidence limits only intended for descriptive purposes. Isotopic values were corrected for the geographic variability using the equations in Table 2.2. Open symbols indicate outliers (>1.5 times the interquartile range) not used to fit the curves. -4 0 4 8 12 16 0 50 100 150 200 250 300 15NF. vesiculosus Population (x103inhabitants) a -4 0 4 8 12 16 0 50 100 150 200 250 300 15NA. nodosum Population (x103inhabitants) b this region (Rio Barja and Rodríguez Lestegás 1996) the influence of riverine nitrate can be considered only of local importance, as reported in other studies (Gago et al. 2005, Bode et al. 2011b). This is supported by our δ15N measurements in nitrate, the first reported for this region, with values close to 5‰ in most cases and particularly in seawater. These values agree with the range reported for subsurface nitrate in the N Atlantic (Liu and Kaplan 1989), while the largest values (>10‰) suggest local influence of nitrate from nitrification of ammonium (Mariotti et al. 1981). 41 Chapter 2: Geographic and anthropogenic variability Systematic observations of coastal waters revealed the importance of local, shortterm upwelling for nutrient inputs in the study area (Álvarez-Salgado et al. 1997, Casas et al. 1997, Nogueira et al. 1998). Because of this nutrient variability, instantaneous nitrogen concentrations and isotopic composition of water samples are not directly reflected in macroalgae collected in the field, in contrast to the findings in laboratory 4949 Ecology of Fucus vesiculosus (Phaeophyceae) at its southern limit of distribution: Growth and production of the early stages of development* Abstract Growth and survival of two populations of Fucus vesiculosus were studied at the southern limit of distribution of this species at the Eastern Atlantic Coast. Experimentally denudated areas at an estuarine and a semi-exposed site in an upwelling area (NW Spain) were followed for 17 months. Three different cohorts were detected during the sampling period. Differences among the three cohorts in terms of growth, reproduction and survival were detected; these differences may be due to the different time of appearance of the different cohorts or the presence of a coverage of previously implanted individuals when the second and third cohorts were recruited. Although the growth of the cohorts recruiting in autumn was higher than for the latter cohorts; the individual growth was represented in all cases by a logistic function, as the fastest rates of increase in length occurred during the first six months of life and maximum length was attained after the thallus reached the first year. In the same way, production was maximum for the first cohort, recruiting in autumn, even when it had the lowest survival rate, because of the rapid growth of survivors during spring and summer. For both populations, reproduction was continuous through the year but it was maximum during spring and summer. Protection from waves might have favoured higher production and standing stock biomass values at the estuarine site than at the semi-exposed site, while turnover rates of biomass were higher in the latter. Contrary to these expectations, most of the nutrients available for the studied populations were not related to upwelling. Despite the fast initial growth of new recruits, both populations appeared to be very sensitive to denudation. keywords: macroalga demography growth reproduction production upwelling Chapter 3: Ecology of Fucus vesiculosus * Viana IG, Fernández C, Bode A (in review) Ecology of Fucus vesiculosus (Phaeophyceae) at its southern limit of distribution: Growth and production of the early stages of development. Eur J Phycol Introduction The shifting of environmental variables in the course of ongoing global warming is expected to impact the performance and distribution of numerous species in marine coastal systems (Wahl et al. 2011). Recent studies have predicted a potential northward shift of intertidal canopy-forming macroalgae along temperate NorthAtlantic rocky shores, especially in the warm-temperate East-Atlantic region from Portugal to Brittany (Jueterbock et al. 2013). As temperature profoundly influences the survival, recruitment, growth and reproduction of seaweeds (Breeman 1988); the study of the response of populations living on the edge could be very important in terms of changes in species’ distribution. Fucoids are the most representative species covering intertidal rocky shores along the European Atlantic coasts from Iceland to Portugal. At their marginal limit of distribution, the NW Iberian Peninsula, these brown seaweeds show a discontinuous distribution, reappearing in isolated patches related with cold water from the springsummer upwelling (Lüning 1990). Therefore, these marginal populations have been shown to be often smaller and more fragmented than central populations. As these populations are considered to live under suboptimal conditions, little variations in their environment could be critical. An increase of seawater temperature, a decrease in the upwelling intensity and other factors like biological interactions or physiological tolerances (Lamela-Silvarrey et al. 2012, NiCastro et al. 2013, Araújo et al. 2014) can affect the performance of the individuals and increase the risk of disappearance of local populations (Jueterbock et al. 2013). Fucus vesiculosus Linnaeus is one of the most common fucoid species and it is usually dominant in the mid intertidal rocky shores at both sides of the North Atlantic Ocean. F. vesiculosus is found in a wide range of wave exposures, from sheltered to moderately exposed areas (Bárbara et al. 1995) and tolerates a large range of salinities (Kautsky et al. 1992). This species has been widely studied but most investigations were restricted to central populations (Keser and Larson 1984a, Carlson 1991, Chapman 1995, Lehvo et al. 2001, Lamote and Johnson 2008, Wahl et al. 2011). In marginal areas, previous studies of Fucus species were mainly focused on community structure and dynamics (Niell 1977a, Fernández and Niell 1982, Bárbara et al. 1995, Lamela-Silvarrey et al. 2012) or on the morphological plasticity of the species (Seoane-Camba 1966, Cairrão et al. 2009, Araújo et al. 2011). However, few studies focusing on growth, production Chapter 3: Ecology of Fucus vesiculosus 51 52 or recruitment have been made in these environments (Niell 1977a, Fuentes 1986, Lamela-Silvarrey et al. 2012, Araújo et al. 2014). Recent studies on genetic variability in this species at its southern limit of distribution suggest population´s different responses in their phenology based on the adaptation to changing habitats and stress tolerance (Billard et al. 2010, Zardi et al. 2013, Jueterbock et al. 2014). The objective of the present study is to quantify growth rates, survivorship, reproduction and production of F. vesiculosus at an estuarine and at a semi-exposed site at Ría de A Coruña (Galicia, NW Spain). These sites are in an upwelling area at the southern limit of distribution of this species. Material and Methods Study sites The Ría de A Coruña is 6 km long and 3 km wide and can be divided in a large bay and a small estuarine zone (Cabanas et al. 1987). The bay has a large oceanic influence and has a mean depth of 25 m. The estuarine zone (Ría do Burgo) has a mean depth of 10 m and a sharp salinity gradient due to the discharge of the river Mero, with a mean flow of 204 hm3 yr-1. The eastern margin of the ria and the estuarine zone are heavily populated (ca. 240,000 inhabitants) while the northern and western margins are characterized by mostly rural landscapes. F. vesiculosus is well distributed in the rocky intertidal areas of this ria from semi-exposed to wave protected areas (Bárbara et al. 1995). The study was conducted at two sites representative of the range of habitats of F. vesiculosus in the region. Mera (43°22’N, 8°20’W) is a rocky semi-exposed shore near the outer limit of the bay where F. vesiculosus is the dominant macroalga from the mid to the lower intertidal, although the population shows a patchy distribution. Individuals in this area typically lack of air bladders and have been described as F. vesiculosus var. evesiculosus (Bárbara et al. 1995). The site at Ría do Burgo (43°20’N, 8° 22’W) is located at the sheltered part of the ria where a dense F. vesiculosus belt is restricted by the presence of a dense population of Ascophyllum nodosum in the upper intertidal zone and by the absence of rocky substrata in the lower zone. The study period lasted 26 months, starting in November 2010 until December 2012. During a 15-month period (November 2010-January 2012). Monthly or 53 bimonthly visits to both sites were made to record the growth, density, biomass and reproduction of F. vesiculosus , along with some water variables. After this first period, bimonthly visits were made just to record the growth of selected individuals. At each visit, salinity (±0.1, Practical Salinity Scale) and temperature (±0.1 °C) of surface water were measured in situ with a portable conductivity meter (YSI Model 30). Samples of surface water were also collected for further determination of dissolved nutrients (NO3 -+NO2 -, NH4 + and PO4 3-) in the laboratory following Grashoff et al. (1983). Intensity of upwelling in the study area was represented by an upwelling index (Lavín et al. 1991) that estimates the Ekman transport of surface water in m3 s-1 by km of coastline computed from geostrophic winds. We used monthly means of the values of the upwelling index data obtained from the Instituto Español de Oceanografía (http://www.indicedeafloramiento.ieo.es) in a grid of 1°•1° centred at 43°N, 11°W. Positive values of this index indicate upwelling of deep waters near the coast while negative values indicate accumulation of shelf surface waters towards the coast (downwelling). Size distributions, growth and reproduction At each sampling site, three 50x50 cm experimental quadrats were randomly set up in the F. vesiculosus dominant zone. The position of the experimental quadrats relative to the Lowest Astronomic Tide (L.A.T.) was between 1.62 and 2.30 m at Ría do Burgo, and between 1.25 and 1.56 m at Mera. The experimental quadrats were denudated in October 2010. All macroalgae constituting the initial undisturbed population inside the quadrats were removed as close to the ground as possible and the substrate was cleaned with a metal brush to ensure no small individuals or the holdfast of any adult individual could remain attached. The material obtained was transported to the laboratory in plastic bags and used for the description of the initial undisturbed population. Observations of the accompanying flora along the studied period were also recorded. All removed individuals were measured, and fronds from the same holdfast were considered as an individual. The total biomass of F. vesiculosus and accompanying flora was determined as wet and dry weight (±1 g). During a 15-month period (November 2010 to January 2012) all individuals within each experimental quadrat were measured monthly, excepting in February Chapter 3: Ecology of Fucus vesiculosus 54 and December 2011. All individuals inside the quadrats were mapped and measured to the nearest mm from the base of the holdfast to the tip of the longest frond. During summer sampling surveys, when the abundance in some of the experimental areas was high, individuals shorter than 0.5 cm were counted and 90 of these individuals were measured to determine their mean length. All individuals were classified in size classes 5-mm wide, from <0.5 to >30.5 cm. Density estimations were made by combining all the results obtained in the three experimental quadrats. Five months after the beginning of the experiment (March 2011), 10 randomly selected individuals within experimental quadrats of each site, and that had been already mapped and measured during the previous period, were labelled. Their growth in length was monitored every month until January 2012, and every 2-3 months until December 2012. At each site, monthly frequency distributions of size classes were calculated from the number of individuals in each size class. Individuals that were first detected in each experimental quadrat were considered as new recruits, and each set of new recruits was monitored as a new cohort. Cohorts were assumed to have normal or log-normal frequency distributions and selection of cohort size ranges was made from the comparison of frequency distributions of consecutive dates. The growth was monitored by monthly changes of the modal length of each cohort fitted to a logistic equation (Niell 1979): Lt=Lmax/(1+e(a1+a2 Age)) where Lt is the modal length at time t , Lmax is the asymptotic length, and a1 and a2 are constants. GraphPad Prism 4 software was used to estimate the fitted curves. To estimate the age of maturity and reproductive periods of F. vesiculosus in the area, the presence of reproductive tips in individuals of the different cohorts within the experimental quadrats was also recorded during the first 15 months of the sampling period (November 2010-January 2012). Demography and production The fate of individuals in the quadrats was monitored based on the monthly maps. New recruits were considered when they were first detected and any individual 55 that disappeared from the experimental quadrats was considered dead. The survivorship (S) of each cohort was estimated from the date when the maximum abundance (maximum recruitment) was detected. Survivorship was calculated as the fraction of individuals remaining from the cohort maximum recruitment and fitted to an exponential decay function with age: S = S0 e-(m Age) where S0 is the density when the cohort was detected (maximum recruitment) and m is the mortality rate. The dry weight biomass of each individual (w, g) was determined using a lengthweight relationship computed from individuals of different lengths (L, cm) and without receptacles sampled at both sites (L=15.460 w0.407, r2=0.938, P<0.001, n=60). Cohort biomass was computed as the sum of the biomass of the individuals recorded in all experimental quadrats and reported as g m-2. The production of each cohort was calculated by the Allen-curve method (Niell 1979, Cousens 1984). This is a graphical method that relates the number of individuals of a cohort (N) with their mean individual weight (w) at different times. After the maximum of abundance is reached (maximum recruitment), only mortality (a decline in density) and individual growth (an increase in mass) occur through the rest of the life cycle of each cohort. The standing stock biomass (B) of each cohort at a given time is defined by N • w under the curve, while the production (P) of the cohort can be computed as the integral under the curve. Standing stock, production and production to biomass ratio (P:B) were computed for each cohort and for the total population for different time intervals. Results Environmental variability The upwelling dynamics in the study area were characterized by a period of positive values between March and August and negative values in autumn and winter months, except for December 2010 when the mean value was also positive (Fig. 3.1). Contrary to expectations, the average upwelling conditions had no effect on the properties of surface seawater when grouped by upwelling and downwelling Chapter 3: Ecology of Fucus vesiculosus 56 periods (Table 3.1). Besides, there were no significant differences between locations for any of the measured variables, and there were only significant differences in temperature and NO3 -+NO2 - between upwelling and downwelling periods (two way ANOVA for the effect of location and period, as fixed factors, and their interaction, Table 3.1). However, these differences are the opposite of those expected from the effect of upwelling, as average values of both temperature and NO3 -+NO2 - during upwelling conditions were lower than during downwelling conditions, suggesting a major role of continental water inputs at both locations. Description of the macroalgal assemblages The mean (±se) biomass of the initial undisturbed population was higher at Mera (643±514 g m-2) than at Ría do Burgo site (415±97 g m-2). Accordingly, size class distributions of the initial population showed a higher number of individuals at Mera, with a predominance of individuals longer than 30.5 cm (top left, Figs. 3.2 and 3.3). The mean (±se) biomass of all accompanying flora summed up to 54±47 and 45±21 g m-2 at Ría do Burgo and Mera respectively. Ulva sp. was quite abundant at both sites at the denudation time (October 2010). In Ría do Burgo, some individuals of A. nodosum were also present at the experimental quadrats, contributing to ca. 10% of total biomass. Ría do Burgo Mera Variable period mean se mean se P t upwelling 15.0 1.3 17.3 1.1 * downwelling 14.3 1.0 18.2 0.8 S upwelling 32.7 0.5 34.1 0.8 n.s. downwelling 28.0 3.9 33.7 1.5 NO3-+NO2upwelling 18.24 5.63 7.74 2.43 * downwelling 38.98 13.08 14.85 2.74 NH4+ upwelling 11.52 3.16 8.90 2.75 n.s. downwelling 21.22 4.70 19.47 3.02 PO43upwelling 1.74 0.27 1.92 0.26 n.s. downwelling 2.18 0.62 2.16 0.42 Table 3.1. Mean and se values of temperature (t, °C), salinity (S), total nitrate (NO3 -+NO2 - , µM), ammonium (NH4 +, µM), and phosphate (PO4 3-, µM) measured at each site and grouped for periods of upwelling (n=7) and downwelling (n=6) following the mean values in Fig. 3.1. P: significance of differences between periods (two way ANOVA, *: P<0.05). 57 During the sampling period, the accompanying flora reached maximum abundance in June 2011. Mera showed the highest species diversity, with Corallina elongata , Osmundea pinatifida , Chondracanthus acicularis , Gelidium pusillum , Cladostephus spongiosus , Leathesia difformis , and some species from the Order Ceramiales and Ulva species, especially U. compressa that were abundant from May until July. In contrast, only G. pusillum , Caulacanthus ustulatus and Chaetomorpha aerea were recorded at Ría do Burgo. Size distributions After the initial denudation of the experimental quadrats three cohorts were identified during the sampling period at each site (Figs. 3.2 and 3.3). However, the progress of the size class distributions of these cohorts was different at both sites. Ría do Burgo was characterized by its fast recovery after scrapping, while recovery at Mera was slower. The cohorts appeared in November 2010, and in March and June 2011 in Ría do Burgo, although new recruits had progressively joined up the population in previous months (Fig. 3.2). In Mera, the cohorts appeared in November 2010 and January and July 2011 (Fig. 3.3). Total and cohort abundance were always higher at Ría do Burgo than at Mera. The lowest recruitment was observed for the first cohort (November 2010) at Ría do Burgo and for the second cohort (January 2011) at Mera, while the highest recruitment was recorded for the third cohort in June and July 2011 at both sites respectively. -1500 -1000 -500 0 500 1000 N D J F M A M J J A S O N D J Upwelling Index (m3s-1 km-1) 2010 2011 2012 Figure 3.1. Monthly means (±se) of the upwelling index (m3 s-1 km-1) computed in a grid of 1°•1° centred at 43°N, 11°W from November 2010 to January 2012. Chapter 3: Ecology of Fucus vesiculosus The growth in length of the studied populations followed a logistic pattern, as it was described for other Fucales (e.g. Niell 1979) but never before for this species. The growth rate reached a maximum when individuals were between 6 months and 1 year old, but stabilized thereafter. The length of old individuals can even decrease due to breakages after gamete release or apical damage and it has been observed that individuals of F. vesiculosus can easily regenerate new fronds from the holdfast after destructive events (Åberg 1989, Malm and Kautsky 2004). These results agree with the reports of significant differences in growth rates between size classes of F. vesiculosus (Fuentes 1986) and other Fucus species (Ang 1991b), with maximum rates in the smaller sizes. 600 250 Ría do Burgo Mera 200 400 Cohort 1 50 100 150 200 Cohort 1 0 0.00 1.78 2.53 2.5 4 400 600 Cohort 2 0 0.00 0.58 1.02 1.0 2 1.01 200 250 Cohort 2 (ind. m-2) 0 200 400 0 00 0 40 0 92 0 94 0 50 100 150 0 00 0 27 0 80 0 83 Density 100 120 100 120 0 . 00 0 . 40 0 . 92 0 . 94 400 600 Cohort 3 0 . 00 0 . 27 0 . 80 0 . 83 100 150 200 250 Cohort 3 0 20 40 60 80 0 20 40 60 80 0 200 0.00 0.13 0.19 0.1 9 0.18 0 50 100 0.00 0.04 0.40 0.5 0 Weight at age (g) 0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Age (yr) 11 Figure 3.7. Allen-curves computed for each cohort at the sampling sites. Density of the individuals at each time was calculated using the survivorship equation in Fig. 3.6 and the individual weight at age (g) obtained from growth equations in Fig. 3.4 and the length-weight relationship. The age (years) corresponding to each weight was represented in a supplementary X-axis. Notice the different density scales for each site and the different weight scales for each cohort. 64 However, the direct comparison of growth rates from different studies is difficult because only maximum or average increases in length per unit of time, but without reference to the age or initial length of the individuals, are reported (Knight and Parke 1950, Fuentes 1986, Bonsdorff and Nelson 1996). From our results, the growth in length expressed as a function of age is markedly nonlinear, implying that growth rates vary continuously from quasi exponential increases during the first months of colonization to almost no increase after the first year. Still, some comparisons can be made by assuming that maximum growth rates reported in other studies are similar to the exponential growth increases averaged over the first year of life in this study. We found increases of up to 2 cm mo-1 that are close to those reported in other locations in Galicia (Fuentes 1986) while values of up to 3 cm mo-1 were reported for French (Lemoine 1913) and British Isles populations (Knight and Parke 1950). Differences among cohorts were also clear in terms of growth rates at both sites. Maximum growth rates were observed for the first cohort (1.8 and 1.3 cm mo-1 in Ría do Burgo and Mera respectively) while individuals from the third cohort were Ría do Burgo Cohort 1 Cohort 2 Cohort 3 Age (yr)B P P:B B P P:B B P P:B 0.0-0.5 37.6439.93 1.06 8.4910.061.194.05 5.591.38 0.5-1.0 211.27 204.730.97 52.8462.82 1.19 8.12 11.81 1.45 1.0-2.0 179.87 76.280.42 32.63 46.64 1.431.30 2.291.76 2.0-3.0 109.460.54 0.01 9.07 0.490.05 0.15 0.01 0.03 0.0 - 3.0 - 321.49 - - 120.01 - - 19.69 - Mera Cohort 1 Cohort 2 Cohort 3 Age (yr) B P P:B B P P:B B P P:B 0.0-0.5 1.57 1.88 1.190.52 0.591.14 0.32 0.40 1.26 0.5-1.0 8.36 10.561.264.77 5.35 1.12 2.17 2.561.18 1.0-2.0 3.194.21 1.32 4.99 6.461.294.63 8.65 1.87 2.0-3.0 0.72 0.02 0.02 1.89 0.12 0.061.42 0.77 0.54 0.0-3.0 -16.66 - - 12.62- - 12.37Table 3.2. Standing stock biomass (B, g dry weight m-2), production (P, g dry weight m-2 period-1), and P:B ratio (period-1) computed for the three cohorts observed at each site for different age periods (years). Chapter 3: Ecology of Fucus vesiculosus 65 the ones with the lowest growth rates (0.6 cm mo-1 at both sites, Fig. 3.4). Based on previous studies, several factors might influence these differences. The time when the different cohorts initially grew vary. Individuals from the first cohort started their exponential growth at spring time, one of the periods of maximum growth of this species at this latitude (Fuentes 1986, Lamela-Silvarrey et al. 2012). On the contrary, individuals from the third cohort started their exponential growth phase in autumn, when conditions for growth are suboptimal. Besides, the presence of a denser and longer canopy of individuals from the previously settled cohorts would have reduced the availability of nutrients and light for smaller individuals of the second and third cohort due to intraspecific competition. This is supported by the absence of any new cohort after the first 9 months of the study, when the three cohorts were observed. In contrast to what it was observed with growth of germlings, the presence of a canopy from the previously settled cohorts seemed to be a good environment soon after settlement, as maximum recruitment was observed in the third cohort at both sites (Figs. 3.2 and 3.3). In this case, the presence of longer individuals might be beneficial for zygote implantation, as this coverage protects the individuals from desiccation, grazers or sedimentation, the major factors that influence survivorship during this first stage (Vadas et al. 1992). The absence of this initial coverage might also be related with the high mortality observed in the cohort 1 at Ría do Burgo (Fig. 3.6). Differences between sites in terms of recruitment and growth rates were also found for all cohorts in our study, although results may be taken with caution due to all factors influencing both sites, apart from the ones considered. As seawater conditions were not very different among sites (Table 3.1), other reasons as the differential wave exposure or the spatial distribution of the populations may be partly responsible of the differences between sites. The population in the scraped area at the semi-exposed conditions of Mera was not protected by other macroalgal species due to the patchy distribution of this species in this location; and individuals from the first cohort were barely 5 cm long when the second appeared (Fig. 3.3). In contrast, the protected conditions at Ría do Burgo would facilitate the implantation and growth of new recruits as the experimental quadrats were surrounded by a dense population of Fucus but also by other large macroalgae (e.g. A. nodosum ). This hypothesis might be supported by the differences found in growth rates related to the degree of wave exposure in other studies, with the result of maximum growth rates and sizes in estuaries, and minimum rates at open ocean locations (Knight and Parke 1950, Kalvas and Kautsky 1993, Bonsdorff and Nelson 1996). However, this is just a hypothesis as other authors 66 have observed that individuals at oceanic influenced sites were longer than individuals at estuarine locations sampled in the same ria although no mention is made to the growth rates or exposure conditions in these studies (Pazó and Romarís 1979). Life span and reproduction A large number of individuals of F. vesiculosus were able to survive through the 17 months of the present study and healthy populations remained in the experimental quadrats (Figs. 3.2 and 3.3). These results agree with the life span of 2-3 yr reported in many studies for most Fucus species, including F. distichus (Sideman and Mathieson 1983), F. vesiculosus (Knight and Parke 1950), F. spiralis (Niemeck and Mathieson 1976) and F. serratus (Knight and Parke 1950). The studied adult populations might have a large reproductive potential (i.e. zygote production, Ang 1991a) as recruitment of the denudated surfaces occurred only one month after removal of adult plants. The identification of different cohorts through the year suggests continuous release of gametes, or at least that this release lasts for a long period of time, as observed in other Fucus populations (Knight and Parke 1950, Keser and Larson 1984a). The reproduction of individuals growing within the experimental quadrats also appeared as a continuous process, as individuals with receptacles were always observed and its proportion in the populations increased through the period of observation (Ría do Burgo) or at least until the first autumn (Mera). In contrast, marked bimodal patterns, i.e. two distinct reproductive peaks, along the year were described for F. vesiculosus from other locations (Knight and Parke 1950, Niemeck and Mathieson 1976, Carlson 1991, Berger et al. 2001). Indeed, a population studied in the years 1983-1984, located only 10 km from Mera in the nearby Ría de Ares, was reported to produce receptacles between autumn and early spring and to release most gametes in early summer (Fuentes 1986). Nevertheless, the presence of fertile individuals, as the ones with receptacles within the quadrats, would not necessarily imply that those receptacles are mature (Berger et al. 2001). Moreover, receptacles could appear in the same individual but in different tips, giving the impression of continuous reproduction at the level of the whole plant (Ang 1992). For most Fucus species, receptacles were first observed in individuals from 4.5 to up to 20 cm long or from 7 months to up to 2 years old (Knight and Parke 1950, Chapter 3: Ecology of Fucus vesiculosus 67 Niemeck and Mathieson 1976, Ang 1991b). Such a large variability might be due to differences among growth rates in different sites, but cannot be directly related to the age of the individuals. Fuentes (1986) observed than the mean length of fertile individuals in the nearby Ría de Ares was 11 cm, although he also found receptacles in individuals 7 cm long, as in the present study. The results provided by our study could suggest that either local factors (like nutrient availability and wave exposure) or decadal fluctuations in the regional oceanography influenced the length of the reproductive period of this species. Even when the annual range of water temperature values reported in both Fuentes (1986) and our study are coincident (13-21 °C) the intensity of the Galician upwelling has decreased significantly since early 1980s (Bode et al. 2011a). A decrease in upwelling implies a decrease in the supply of new nutrients to the surface, but the concentrations of most nutrients near A Coruña have not shown any significant trend in the last two decades (Bode et al. 2011a), suggesting a compensation of the loss of upwelling nutrients with an enhanced input of terrestrial (mostly anthropogenic) nutrients. Besides, the variables measured in the surface water do not support a major role of upwelling dynamics in the present study. Runoff is maximum during autumn and winter, as indicated by the low salinity values in Table 3.1, and may introduce significant amounts of nutrients from terrestrial sources. Therefore the long period of reproduction found in this study cannot be attributed to a general increase in nutrients, at least at a regional scale. Ladah et al. (2003) reported that protection from waves neither explains reproductive success nor influence in fertilization. Therefore the low impact of waves is not the reason for the persistence of receptacles in the studied estuarine population through the year in contrast with the findings of Fuentes (1986) in a more exposed population. Production As expected from growth and recruitment results, the rank in production of the cohorts was led by the first cohort at both studied sites while the third cohort was generally the less productive for all time-periods considered (Table 3.2). Early implantation in fall allowed the increase of the standing stock of the first cohort with rapid individual growth in the spring despite the low recruitment. In contrast, a relatively high recruitment in summer did not warrant high production 68 because of the reduced growth rates and high mortality observed during winter for the second and third cohorts. Estimated total production reached its maximum when the cohort was from 0.5 to 2 years old, depending on the site considered (Table 3.2). At this age, the increment in biomass depends only on the growth of survivors, and the presence of other Fucus individuals results in a competition due to the limitation for light and nutrients. In all cases, but particularly for the first and second cohort, the values of production estimated for Ría do Burgo largely exceeded those for Mera, supporting the hypothesis that growth and survivorship are higher in wave protected, estuarine sites than at exposed, oceanic sites (Knight and Parke 1950). However, another study in Ría de Arousa (Galicia) reported lower production for estuarine sites than for oceanic sites (Fuentes 1986), suggesting that additional factors other than exposure and nutrients are also influencing local production. All cohorts of the population studied at Ría do Burgo showed a lower turnover rate than those at Mera (Table 3.2). A larger turnover of biomass at exposed sites than at estuarine sites was also reported for other populations of F. vesiculosus in Galicia (Fuentes 1986) and was related to the enhanced energy flow in the former, favouring the access to nutrients and light in turbulent environments. For both sites, the production during the first year was almost equivalent to the standing stock biomass (P:B~1) indicating low renovation rates. Mean annual production of the three cohorts in our study was much lower than those reported for Fucus species in other close sites but some decades ago (Table 3.3). The same reduction pattern in the production of fucoid species (including F. vesiculosus ) was observed in the Cantabrian coast over the last 30 years (LamelaSilvarrey et al. 2012). Although no direct causal mechanisms can be inferred to this fact, the general trend of increasing temperature due to climate change could be responsible (Lamela-Silvarrey et al. 2012). Still, comparisons of data from this study might be done with caution as methodological variations may in part explain this discrepancy. Previous production studies only consider average changes in biomass between consecutive sampling periods (generally 1 month apart) and our estimation is based on conservative modal values for growth and demographic parameters. Cousens (1984) has shown that differences in the assumptions when computing production using Allen curves in macroalgae may produce estimates varying by a Chapter 3: Ecology of Fucus vesiculosus 69 factor of 2, and even larger differences may result when considering only part of the annual cycle in the computations (Niell 1977a). Besides, in the present study the production was computed for a newly developed population after denudation, while in other studies the production is estimated for mature populations. The low production rates estimated imply that the studied populations will recover slowly after denudation even when there were old plants remaining in the vicinity, particularly at the wave exposed site where the distribution of F. vesiculosus was patchy. Therefore these populations are sensitive to mechanical damage causing a loss in biomass and production not only in the dominant macroalgal cover but also in the accompanying flora and fauna. Species Production Reference F. spiralis 922.8Niell 1977a F. vesiculosus 979.9 - 1828.5 Fuentes 1986 1431.8 - 2255.01 Lamela-Silvarrey et al. 2012 41.65 Mera, this study 461.79 Ría do Burgo, this study Table 3.3. Production (g dry weight m-2 yr-1) of Fucus spirali s and F. vesiculosus in the literature and this study. 70 7373 Growth and production of new recruits and adult individuals of Ascophyllum nodosum in a non-harvested population at its southern limit (Galicia, NW Spain)* Abstract Populations near the geographic distribution limits of the species are considered to live under suboptimal conditions, and hence slight environmental changes can be critical for their survival. The potential sensitivity to disturbances of the long-living macroalga Ascophyllum nodosum was analyzed by the determination of growth, recruitment, mortality, and production of biomass of a population near its southern distribution limit. Recruitment, survival and growth rates of <2 yr old individuals were determined in a new population growing in experimentally denudated squares. Demographic data for >2 yr old individuals were obtained from individuals in the original population after estimating their age from the number of gas bladders in the thallus. Growth and survival were described as continuous nonlinear functions of age applied to the population and were further used to make demography-based production estimates. Recruitment of A. nodosum in denudated substrates seemed to require a previous cover of other macroalgae (as Fucus vesiculosus ) as the only cohort detected during the 26-month period was observed after the F. vesiculosus individuals started to increase. The low production estimates (2,033 g m-2 for a 10-yr period) and poor recruitment may indicate a slow recuperation of this population to denudation. However, the large variability observed in the estimated growth curves of different populations along this southern distribution area suggests a large influence of local conditions that may help to overcome environmental changes at regional scales. Chapter 4: Ecology of Ascophyllum nodosum keywords: demography growth curve fucoid size distribution gas bladder recruitment * Viana IG, Fernández C, Bode A (in press) Growth and production of new recruits and adult individuals of Ascophyllum nodosum in a non-harvested population at its southern limit (Galicia, NW Spain). Mar Biol, doi: 10.1007/s00227-014-2553-0 population are recorded at one moment in time. Following this method, the growth of A. nodosum was estimated from the original population sampled at Ría do Burgo in October 2010. Temporal and geographic variability in growth was analyzed after estimating growth curves from samples collected at Ría do Burgo in 2006 and at other sites along the coast of Galicia (Fig. 4.1). These estimated curves were also compared with the curves determined for Cíes Islands and San Simón by Niell (1979). To test that the formerly assumptions were correct, the appearance of the annual gas bladder in new recruits and labeled adult individuals was checked. Two approaches were made, i) the percentage of new recruits within the experimental squares with the first gas bladder was recorded; and ii) seven adult individuals close to the experimental squares were labeled to follow the appearance of the annual gas bladder. Demography and production of new recruits from the experimentally denudated squares The fate of individuals within the experimental squares was monitored based on the monthly records of their position within each experimental square. New recruits were considered when they were first detected. Any individual that disappeared from the experimental squares was considered dead. Survivorship (S) was calculated as the fraction of individuals remaining from the cohort maximum recruitment and fitted to an exponential decay function with age: S = S0 e-(m Age) where S0 is the density when the cohort showed maximum recruitment and m the mortality rate. The dry weight biomass of each individual (w, g) was determined using a lengthweight relationship computed from a set of individuals of different lengths (L, cm) and without receptacles collected at Ría do Burgo (L=21.962w0.418, r2=0.962, P<0.001, n=42). The total biomass of the population was computed as the sum of the biomass of the individuals recorded in all experimental quadrats and reported as g m-2. The production was calculated by the Allen-curve method (Niell 1979, Cousens 1984). Production estimates using this technique can be obtained graphically by relating the 80 number of living individuals of a cohort (N) and the mean weight of those survivors (w) at corresponding times (Crisp 1971). After the maximum of abundance is reached (maximum recruitment), only mortality (a decline in density) and individual growth (an increase in biomass) occur through the rest of the life cycle of the cohort. The standing stock biomass (B) of each cohort during a period of time is given by the square defined by N • w under the curve, while the production (P) of the cohort can be computed as the integral under the curve. Standing stock, production and production to biomass ratios (P:B) were computed for different time intervals. Results Environmental settings and characteristics of the experimental squares The site showed large ranges of variation in both temperature and salinity (Fig.4.2). Temperature reached a maximum of ca. 23 °C by the end of summer while minimum values in winter reached 10 °C. Low salinity values were recorded during fall and winter. Nutrient concentrations also varied during the sampling period (Fig. 4.2). Nitrate plus nitrite and phosphate were maximum during winter months, while values in early summer were below 5 µM. The fact that the lowest values were found during upwelling months (spring and summer) and the inverse variation of salinity vs. nitrate plus nitrite suggests a major role of river nutrient inputs in this location. The mean (±se) dry biomass of A. nodosum individuals removed from the original population within experimental squares was 3,456±1,241 g m-2. Among the possible accompanying species, only F. vesiculosus was present within all the experimental squares in the original macroalgal assemblage, averaging 108±144 g m-2. During the first months after denudation several specimens of F. vesiculosus appeared in the experimental squares, reaching higher densities than A. nodosum (Fig. 4.3). Nevertheless, other seasonal macroalgae were recorded during the sampling months, with the maximum number of species detected in June 2011. From May until July, when water temperature was high, Chaetomorpha aerea was present, and among Rhodophyceae, only Gelidium pusillum and Caulacanthus ustulatus were recorded. Size distributions of new recruits from the experimentally denudated squares After the denudation (October 2010), and during the period of study (~2 yr), new A. nodosum recruits were observed all along the sampling period. During the first months the recruitment of this species within the experimental squares was Chapter 4: Ecology of Ascophyllum nodosum 81 0 25 50 75 100 O N D J F M A M J J A S O N D J F Nitrate+nitrite or ammoium (µM) 10 20 30 40 O N D J F M A M J J A S O N D J F Salinity 10 15 20 25 O N D J F M A M J J A S O N D J F Temperature (⁰C) 0 1 2 3 4 5 O N D J F M A M J J A S O N D J F Phosphate (µM) ba c d 2010 2011 2012 2010 2011 2012 Figure 4.2. Variations in water temperature (a, °C), salinity (b), nitrate+nitrite and ammonium (c, µM, continuous and dashed lines respectively), and phosphate (d, µM) at Ría do Burgo site during the first 15 months of the study period. minimum, and only up to 50 individuals m-2 were mapped until April 2011 (Figs. 4.3 and 4.4). In spring, when the number of F. vesiculosus individuals started to increase, the number of individuals of A. nodosum also started to rise, although at slower rates than Fucus (Fig. 4.3). In accordance, and based on the monthly variation of the sizedistributions, the start of the only cohort detected during the study was identified in May 2011, when a marked density peak was observed (Fig. 4.4). 82 Since June 2012 a few A. nodosum individuals longer than 12.5 cm were observed, but their abundance was very low compared to size classes shorter than 5 cm. And by the end of the sampling period (December 2012) the highest densities were still recorded among individuals shorter than 5 cm. Individuals between 12.5 and 26 cm (the maximum length observed) reached approximately 60 indiv m-2, while in the original population most individuals were longer than 60 cm (Fig. 4.4). 0 2000 4000 6000 N D J F M A M J J A S O N D J Density (indiv. m -2 ) A. nodosum F. vesiculosus 2010 2011 2012 Figure 4.3. Variations in density (individuals m-2) of the new recruits of A. nodosum and F. vesiculosus in the experimental squares at Ría do Burgo during the first 15 months of the study period. Growth and appearance of the annual gas bladders in new recruits and adult individuals The first gas bladder in the new recruits within the experimental squares appeared in March 2012 (Fig. 4.5). All individuals within the experimental squares that showed a gas bladder were between 7 and 18 cm long and were not more than 17 months old, the time since the denudation was done. But during the 26 months of the study not all recruits longer than 7 cm had a gas bladder. Among adult labeled individuals, the annual gas bladder was present at the tip of all individuals in the spring of both recorded years. During the first spring (2011) the bladder was developed later (between March and April) than in the second year (January-March). The month after a new gas bladder was observed for the first time, it was not at the tip of the individual anymore, and in some individuals a bifurcation was already present following the formation of the gas bladder. This last feature was not conspicuous to all individuals, as some of them showed 2 vesicles in a row or even a bifurcation on the vesicle. Although there were some individuals within the experimental squares exceeding 12 cm at the end of the study, the modal length of the new recruits was much lower (Fig. 4.4). To complete the estimated growth curve of Ría do Burgo, data from the individuals of the original population within the experimental squares were Chapter 4: Ecology of Ascophyllum nodosum 83 Figure 4.4. Variations in density (individuals m-2) of new recruits of A. nodosum by 0.5-cm size-classes (<12.5 cm) in the experimental squares at Ría do Burgo during the study period. The size distribution of the original population before denudation is shown in the upper left panel. The time after denudation (days) along with the date of observation is also indicated on each histogram. Distributions shown are from April 2011 just before first density peak was detected. Note the different scales of the yaxis. 84 combined along with the modal length of the new recruits (Fig. 4.6a). The estimated growth curve showed slow growth during the first years of life and an exponential growth until the individuals were 95 cm long (Table 4.1). The populations from other localities showed growth curves similar to those obtained at Ría do Burgo (Fig. 4.7a). The growth curve of Cíes Islands showed the slowest growth rate (Fig. 4.7a). There were no differences between the growth curves of Ría do Burgo at different sampling years; and for both years, the oldest individuals observed were 9 years old (Fig. 4.7b). Survivorship and production of the new recruits within experimental squares The survivorship was estimated from the date the maximum abundance of the cohort was detected. The new population displayed high recruitment (~2,500 indiv m-2) but also high mortality among individuals younger than 2 years old (Fig. 4.6b). The estimated survivorship curve was completed with density values of older individuals (≥4 years) of the original population. The mortality among individuals older than 3 years old was estimated to be much lower. 0 20 40 60 80 100 N D J F M A M J J A S O N D J F M A M J J A S O N D % individuals with respect to the total individuals > 7 cm Without gas bladder With gas bladder 2010 2011 2012 ****** ** Figure 4.5. Record of the appearance of the annual gas bladders in new recruits longer than 7 cm and percentage of those new recruits with or without a gas bladder within the experimental squares at Ría do Burgo during the sampling period. *: No data. Chapter 4: Ecology of Ascophyllum nodosum 85 Fig. 6 1 2 4 8 0 2 4 6 8 10 Density (indiv. m -2 ) Age (yr) ln S = 7.84 - 0.72 Age 0 50 100 150 0 2 4 6 8 10 Length (cm) Age (yr) 0.0 3.2 28.3 Weight (g) a b 0 1000 2000 3000 0246810 Density (indiv. m-2) Age (yr) c Figure 4.6. Growth in length (a, cm), ln survivorship (b, individuals m-2) and Allen diagram (c) of A. nodosum population at Ría do Burgo. The open dots in panel a indicate adult individuals measured in the original population (2010) dated using the number of gas bladders and length. The black dots in panel a indicate the modal length of new recruits observed during the sampling period (Fig. 4.4). The parameters of the estimated growth curve in panel a are listed in Table 4.1. In panel b, the black dots indicate the density of new recruits observed during the sampling period (Fig. 4.4), the open dots are density estimations of the new recruits not used in the computation of the survivorship curve and the gray dots represent the density estimations of adult individuals from the original population (2010). 86 Despite the low recruitment, individuals between 2 and 6 years old were the most productive (Table 4.2). This is due to the higher standing stock biomass of the individuals at these ages because of the higher growth rates (Fig. 4.6a). Individuals between 8 and 10 years old may suffer several breakages, diminishing the standing stock biomass of the aged population, and in consequence reducing its production. The P:B ratio also supports that the biomass produced annually is over the standing stock (P:B>1) until the population is between 6 and 8 years old. Chapter 4: Ecology of Ascophyllum nodosum 87 Age B P P:B (yr) (g m -2 ) (g m -2 period -1 ) (period -1 ) 0-133.52 46.95 1.40 1-296.42 104.161.08 2-4 455.99 708.66 1.55 4-6 568.68 925.00 1.63 6-8 231.04 229.75 0.99 8-10 62.1918.430.30 0-10 2032.96 Table 4.2. Standing stock biomass (B, g m-2), production (P, g m-2 period-1) and P:B ratio (period-1) estimated for the population of A. nodosum at Ría do Burgo using the Allen method for different time periods (yr). Sampling year Latitude N Longitude W L max a1 a2 n r 2 Foz 2010 43.56468 - 7.24599 135.90 3.61 - 0.55 42 0.88 Cabanas 2010 43.41146 - 8.17255 113.50 3.21 - 0.58 46 0.79 Ría do Burgo 2010 43.32777 - 8.07357 95.09 3.70 - 0.80 133 0.77 Ría do Burgo 2006 43.32777 - 8.07357 101.40 3.21 - 0.61 37 0.80 Camariñas 2007 43.13694 - 9.17705 97.07 3.40 - 0.90 25 0.86 San Simón * 1978 42.30645 - 8.62840 100.00 2.27 - 0.43 - 0.82 Cíes Islands * 1976 42.22461 - 8.90689 105.65 2.40 - 0.18 - 0.90 * from Niell (1979) Table 4.1. Parameters of the logistic equations describing growth in length (cm) of A. nodosum computed for different sites along the Galician coast (see Figs. 4.1 and 4.7). The equation is in the form: Length = Lmax / (1+ e[a1+a2 Age]), where Lmax is the asymptotic length (cm) and age is measured in years. a1 and a2 are constants; n: number of data; r2: determination coefficient. Discussion The studied population of A. nodosum showed low recruitment, growth, survival and productivity rates. These results are consistent with low recovery capacity associated to environmental stress at the southern limit of distribution of this species (Araújo et al. 2014). However, the local variability observed in the estimated growth curves suggests the existence of different local life history traits that may enhance the survival of this species under stress. Recruitment The slow recovery of A. nodosum following natural or experimental denudation was previously reported in natural and harvested areas (Knight and Parke 1950, Baardseth 1970, Vadas et al. 1990, Jenkins et al. 1999, Cervin et al. 2005, Ingólfsson and Hawkins 2008). At the experimental squares of Ría do Burgo, new A. nodosum recruits were almost absent during the first six months of the study (Figs. 4.3 and 4.4) when most individuals outside the squares did not have receptacles (pers. obs.). The start of the cohort was established in May 2011, when the first peak (mode) was observed, in agreement with the presence of mature receptacles in March that were reabsorbed in April after releasing the gametes. Nevertheless, increasing densities were observed the subsequent months after the cohort was established. The existence of a bank of gametes that gradually developed could have originated the subsequent increase in density of the cohort months later it was established. However, recruitment may not be simply related with the reproduction of adult individuals. For instance, a peak in recruitment was not observed in spring 2012 (Fig. 4.4), likely because the already high density of plants within the experimental squares. They may have prevented the settlement of significant numbers of zygotes or at least their survival until they were first detected because of the intraspecific competition for space and resources. Within the experimental squares, the rock was initially colonized by F. vesiculosus individuals, even though the biomass of this species was 32 times lower than A. nodosum biomass in the original population. The initial appearance of Fucus spp. individuals in denudated areas was also observed in northern areas (Keser and Larson 1984b, Jenkins et al. 1999, Dudgeon and Petraitis 2001, Cervin et al. 2005). Soon after settlement, a dense canopy seems to be a good environment that enhances survivorship of several macroalgal germlings (Cervin et al. 2005, Choi and Norton 88 0 50 100 150 Length (cm) Foz Cabanas Camariñas San Simón Cies Islands 0 50 100 150 0 5 10 15 20 Length (cm) Age (yr) Ría do Burgo 2010 Ría do Burgo 2006 a b Figure 4.7. Estimated growth in length (cm) of populations of A. nodosum at several locations in Galicia (a) and at Ría do Burgo for years 2006 and 2010 (b). The parameters of the logistic equations are given in Table 4.1. The curves of San Simón and Cíes Islands are from Niell (1979). Chapter 4: Ecology of Ascophyllum nodosum 89 2005, Sánchez and Fernández 2006), although later on, high densities may derive in a inhibition of growth rates due to the competition for nutrients, light, space, wave action or due to sweeping by large individuals (Vadas et al. 1992, Creed et al. 1996, Viejo et al. 1999, Steen and Rueness 2004, Cervin et al. 2005, Choi and Norton 2005). Densities of new recruits observed in the present study (up to 2,500 indiv m-2) were much lower than those reported for central populations after denudation, Introduction Concern with coastal eutrophication has increased in the last decades due to increasing N loading associated with the growing human population in these areas. Numerous methods have been developed to identify and quantify the N sources. Among them, the ratio of the stable isotopes of N (δ15N) in macroalgal tissues has been increasingly used. The δ15N not only allows detecting the presence of anthropogenic N that is actually available for macroalgae in coastal water, but also allows estimating the intensity of the effluents and detect disturbances before alteration in structure and function occur in the ecosystem (McClelland et al. 1997, McClelland and Valiela 1998a, 1998b; Costanzo et al. 2001, Gartner et al. 2002, García-Sanz et al. 2010, 2011; Carballeira et al. 2013). Among macroalgae, Phaeophyceae have been widely used for monitoring loads of N and other substances (e.g. heavy metals) (Viana et al. 2010, 2011). They are good biomonitors because they are perennial, resistant to pollution and accumulate dissolved substances in their tissues (Phillips and Segar 1986). They take up, assimilate and accumulate N in excess to growth demands so they can be used as integrators of N availability (Hanisak 1983), offering time-integrated measures of the exposure time. The Fucaceae Fucus vesiculosus and Ascophyllum nodosum are two long-lived canopy-forming macroalgae widely distributed at both sides of the Atlantic Ocean. They can live up to 15 yr in the case of A. nodosum (Niell 1979). Both of them show apical growth, so if growth rates are known (Viana et al. in review b, in press), different segments along the frond can be related with particular environmental or water conditions (Savage and Elmgren 2004, Raimonet et al. 2013, Carballeira et al. 2014). Moreover, A. nodosum fronds develop a gas bladder in the tip that generally occurs once a year (David 1943, Viana et al. in press). This annual bladder enables to estimate the minimum age of an individual and delimitate its annual growth (Niell 1979, Viana et al. in press). Both species of macroalgae have been previously used to monitor N sources (Hobbie et al. 1990, Savage and Elmgren 2004, Deutsch and Voss 2006, Bode et al. 2011b, 2014; Viana et al. 2011, Raimonet et al. 2013, Viana and Bode 2013, Carballeira et al. 2014). The use of stable isotopes on macroalgae to detect N loadings requires some assumptions related to their ecology and physiology. For instance, the accumulation of heavy isotopes depends on the preference of light isotopes (or fractionation) 97 Chapter 5: Experimental assessment of the macroalgal δ15N during uptake, and on the uptake capacity of tissues. Concentration dependent fractionation during uptake has been reported for some algae, as diatoms (Wada and Hattori 1978, Pennock et al. 1996). However, experimental studies in different macroalgal species, as the Chlorophycea Ulva (formerly Enteromorpha ) intestinalis (Cohen and Fong 2005) and Ulva pertusa (Dudley et al. 2010) or the Phaeophycea Cystoseira mediterranea (García-Sanz 2009) demonstrated that, at least those macroalgae, did not exhibit concentration dependent N isotope fractionation. Some studies were based on the relatively long life span of F. vesiculosus, and interpreted the isotopic composition of different sections of their thalli as the result of past pollution events (Savage and Elmgren 2004, Raimonet et al. 2013, Carballeira et al. 2014). The main assumption of these studies is that only the growing tips of the thallus take up nitrogen and, therefore, the isotopic composition of a given section of the thallus would reflect the isotopic composition of the dissolved nitrogen in the surrounding water at the time of growth. However, some questions need to be tested to fully interpret the data obtained in these studies. First, Fucaceae do not have a specific transport tissue, but the pores of the sieve plates should enable a continuous system of cytoplasm for the translocation of materials longitudinally (Moss 1983). There is experimental evidence of long distance transport of organic 14C, 86Rb or 32P (Penot and Penot 1979, Diouris and Floc´h 1984, Raven 2003). If the transport of nitrogen along the thallus also exists, it would directly affect the retrospective identification of past nitrogen sources. Second, most studies assume that the isotopic composition of tissues does not change for at least several months, given that these species generally show low variability in δ15N values at monthly time scales (Gartner et al. 2002, Raimonet et al. 2013) but no data of N-specific uptake and turnover rate were available for this species. For assessing the feasibility of using A. nodosum and F. vesiculosus for the isotopic differentiation of local N sources, two sets of experiments were made under laboratory conditions. The first experiment aimed to determine the equilibration of N isotopes in the growing tips and older parts of the fronds by growing them under water with different N origins. The second experiment aimed to detect nitrogen transport along their thalli and to test if all the parts of the frond have the capacity of taking up NO3 - by using artificially 15N-enriched water. The latter approach also allowed the estimation of N turnover rates in different sections of the thallus. 98 Material and Methods Experiment 1: N isotope equilibration Water samples–The first laboratory experiment was conducted with water from 3 different sites: water from an urbanized watershed, from a forested watershed, and from an oceanic influenced site which was considered the control. The first two sites are Childs River (CR) and Sage Lot Pond (SLP), which are part of the Waquoit Bay National Estuarine Research Reserve, Massachusetts (Fig. 5.1). The Waquoit Bay estuarine system is a complex of sub-estuaries with different N inputs from their watersheds, and thus, with differing ambient N concentration and origin (Valiela et al. 1992, Valiela et al. 1997). The CR estuary (41°34’N, 70°32’W) is surrounded by the most urbanized watershed in the Waquoit Bay system. Nutrients (primarily nitrate) are delivered to the CR estuary from the watershed via groundwater flow (Valiela et al. 1992). In contrast, SLP (41°55’N, 70°50’W) is a forested watershed receiving a low N load, with NH4 + as the dominant dissolved inorganic nitrogen (DIN) form (Valiela et al. 1997) and is surrounded by salt marshes. The control site was placed at Nobska Beach (41°51’N, 70°65’W), which is an oceanic influenced site with no terrestrial or anthropogenic inputs draining in the area (Fig. 5.1a). Experimental design– Individuals of A. nodosum and F. vesiculosus were collected at Quissett Harbor and Nobska Beach respectively, in Woods Hole, Massachusetts (Fig. 5.1a); and they were transported in coolers to laboratory. Macroalgae were kept in tanks with continuous seawater flow (15.7±1.6 °C) and low light intensities during the night (less than 12 hours) until the start of the experiment. A. nodosum individuals of 14.6±2.6 cm long and with 2 or 3 gas bladders, and F. vesiculosus individuals of 10.7±2 cm long were selected to run the experiment. Individuals with visible damage or epiphytes were avoided. Macroalgae (n=4 for A. nodosum , n=3 for F. vesiculosus ) were placed in three different 1 L Erlenmeyer flasks containing CR, SLP or Nobska unfiltered water. The study was run in triplicate with each replicate in a separate flask for each of the three treatments over a period of 22 days for A. nodosum and 12 days for F. vesiculosus . Samples were taken at the start of the experiment (t=0) and at subsequently exponential times. The different time scales for each species were chosen based on the previous knowledge of growth rates of the species. A control 99 Chapter 5: Experimental assessment of the macroalgal δ15N 70°0’W71°0’W 42°10’N 42°10’N 42°0’N 42°0’N 41°50’N 41°50’N 41°40’N 41°40’N 41°30’N 41°30’N 41°20’N 41°20’N ab ab 021 Kilometers 03 1.5 Kilometers 0 5025 Kilometers Quissett Harbor Nobska Beach Sage Lot Pond Childs River Cape Cod MA Figure 5.1. Location of the study sites at Cape Cod, Massachusetts, USA (Basemap: USGS). flask with no macroalgae was established for each water treatment and maintained under the same conditions as the experimental flasks. For comparison with experimental individuals, native individuals of F. vesiculosus were collected along with water samples where present (i.e. CR and SLP) and analyzed for stable isotope composition. Local populations of A. nodosum were not found at the sites selected for water collection. 100 Experiments were carried out in a culture chamber with 18:6 light:dark cycle at light intensities varying between 390-450 µE m-2 s-1 under 18-20 °C air temperature oscillation between night and day respectively. Water aeration was maintained with air pumps and diffusers and water temperature oscillated around 24.08±0.06 °C. Water was replaced every 2 days to avoid nutrient depletion. Samples of water were collected before and after replacement to quantify the variation in DIN concentrations among times and sites and to check macroalgal consumption. Salinity and temperature were measured with a portable conductivity meter (YSI Model 30) every time the water was changed. The macroalgal samples used for δ15N and N and C content were separated with a glass spatula. The growing tip (1 cm) was sampled at all sampling dates during the experiment for both species. Additionally, at the start of the experiment (t=0) and at the endpoint, the growing tip (1 cm) and all intervesicular segments were sampled in A. nodosum individuals, while for F. vesiculosus individuals only the growing tip (1 cm) and the basal segment of the frond were sampled. All macroalgal samples were rinsed with Milli Q water and frozen (-20 °C) before processing. Later, samples were defrosted and dried (50 °C) until constant weight before grinding into a homogeneous powder prior to isotopic and elemental analysis. Macroalgal growth–To measure macroalgal growth response to the different water samples, the wet biomass of each frond was recorded at the beginning of the experiment and at the time the frond was sampled. Individual growth rates (µ) were calculated as a percent increase in biomass per day (% d-1): µ=100[Ln(Nt/N0 )]/ t where Nt is the biomass on day t , N0 is the initial biomass, and t is time in days of incubation (Lobban and Harrison 1994). Nutrient sampling and analysis–Changes in concentration of NO3 -+NO2 -, NH4 +, and PO4 3were determined during the experiment to quantify differences in ambient nutrient concentrations among water samples. Water samples were frozen until analysis of nutrient concentrations. Nitrate and phosphate were determined using standard colorimetric assays in a Lachat Auto Analyzer (Cd Reduction). Ammonium 101 Chapter 5: Experimental assessment of the macroalgal δ15N concentrations were determined by spectrophotometry following the Indophenol method. Detection limit was 0.25 µM for any of the three nitrogen species. Experiment 2: 15N enrichment experiment An enrichment experiment was done to determine N-turnover rates in different sections of the thallus and to test: i) the occurrence of transport of N along the thallus, from the tip to the basal segment of the frond, ii) the occurrence of transport of N from the basal segment of the frond to the tip, and iii) to quantify the uptake rates of the growing tips and mature parts of the thallus. As in the previous experiment, A. nodosum and F. vesiculosus were collected at Quissett Harbor and Nobska Beach respectively (Fig. 5.1a). Macroalgae were transported in coolers to laboratory and maintained under the same pre-incubation conditions as previously described. For these experiments, A. nodosum individuals were 23.2±0.9 cm long and had 4 gas bladders, and F. vesiculosus individuals were 12.7±1.1 cm. The selected individuals did not show apparent damage or epiphytes. Treatment water was created by adding a stock solution of 10mM K15NO3 (99 atom % 15N) to 2 L of a final volume of seawater. The final concentration was ~120 µM, with 98.8% atom % 15N enrichment. Nitrate was selected as the tested nutrient as it is the dominant inorganic nitrogen compound in sewage. To test i) and ii), experiments were divided in two periods: a first 4-h period under the stock solution, followed by a 24-h period under control seawater. During the first period, only the tips (i) or the basal segment of the frond (ii) of three different individuals of each species were submerged, while the non-submerged parts of the thallus were manually vaporized with control seawater at regular intervals (~20 min) to avoid desiccation. Macroalgae were maintained inside the culture chamber under the same light and temperature conditions as in the previous experiment. After this first 4-h period, individuals were gently washed with seawater and transferred individually to an Erlenmeyer flask with 1 L of control seawater. They were kept during 24 hours under the same conditions of temperature, light and aeration as in the previous experiment. After both incubation periods, all individuals were immediately subsampled for stable isotope determinations. Each A. nodosum individual was divided into tip (11.5-cm fragment measured from the distal part) and intervesicular segments. Those of 102 F. vesiculosus were divided into tip (1-cm fragment from the distal part) and regular length segments (~3 cm) from the tip to the base. The lateral vegetative or reproductive branches of A. nodosum or reproductive tips of F. vesiculosus were discarded. To test iii), the uptake capacity of the tip and non-growing parts of the thallus, three individuals of each species were completely submerged in the treatment solution for 2 h. Macroalgae were maintained inside the culture chamber under the light and temperature conditions as in the previous experiment. To exclude the possible transport of inorganic N along the thallus, macroalgae were subsampled immediately after the incubation period. Macroalgae were subsampled following the same procedure as previously described for i) and ii). During each of the three treatments, control individuals of A. nodosum (n=3) and F. vesiculosus (n=3) were maintained in the same conditions as the experimental individuals but in 1L Erlenmeyer flasks with control seawater. Internal nutrient content and δ15N analysis N stable isotope and elemental analyses for N and C content to estimate the tissue C:N were performed for all samples. Aliquots of ca. 2.5 mg of macroalgae samples were used. Samples were placed in tin capsules and introduced into an isotope-ratio mass spectrometer (Thermo Finnigan Mat Delta Plus) via an element analyzer (Carlo Erba CHNSO 1108). Isotopic results are expressed in delta notation: δ15N = [(15Nsample:14Nsample/15Nstd:14Nstd)-1] x 1000 where the standard (std) is atmospheric N2. Precision (se of 5 replicates) was better than 0.05‰ for either IAEA-N-2, IAEA-N-1 or IAEA-NO-3 standards. The coefficient of variation of triplicate sample aliquots was always <2%. Statistical analyses and calculations Comparison of nutrient concentrations among water samples was done by analysis of variance (ANOVA). This test was also used to analyze differences among sites and macroalgal segments along the thallus at the end of the isotope equilibration experiment, and to study differences between macroalgal segments within individuals from the same site. When significant differences were detected, a posteriori StudentNeuman-Keuls (SNK) tests for multiple comparisons were used to detect differences 103 Chapter 5: Experimental assessment of the macroalgal δ15N among groups. Differences in the δ15N and C:N in the growing tips of macroalgae over time were tested using a general linear model (GLM) univariate procedure using the site and time as fixed factors. Experimental samples of the 15N enrichment experiments were compared with the control samples to test the atom % 15N enrichment using a paired-samples t -test, which compares two measurements of the same sample before and after the treatment. All tests were carried out with SPSS Statistical Software. To estimate N uptake in the enrichment experiment we used the N specific uptake rate, which was calculated from appearance of the 15N in the macroalgal tissue: N specific uptake= (atom% 15Nf - atom% 15Ni)/ R ∙ t where atom % 15Nf and atom % 15Ni are the final and initial atom % 15N enrichment of macroalgal thallus respectively, R (%) is the calculated exponential average of the initial and final atom % enrichment of NO3 -, and t is the time in hours. The inverse of the N specific uptake-rate was used to estimate the turnover time ( tr ) in days that would take to renovate the total N of a particular macroalgal fragment. Results Experiment 1: N isotope equilibration rates Concentrations of all inorganic nitrogen compounds during the experiment with A. nodosum in September were higher than those found during the F. vesiculosus experiment in August (Table 5.1). In the former case, water from CR had more nitrate and ammonium than water from the other sites but showed similar phosphate concentrations. In contrast, during the F. vesiculosus experiment, the oceanicinfluenced site (Nobska) showed larger nitrate and lower ammonium and phosphate concentrations than those at the other experimental sites, which showed similar concentrations of all nutrients. In all cases, DIN:PO4 3values were low, indicating potential nitrogen limitation. The macroalgal growth response to nutrient changes differed between species, although the pattern was very similar among sites within the same species (Fig. 5.2, Table 5.2). Overall the growth of A. nodosum was higher than the growth of F. vesiculosus . 104 A. nodosum F. vesiculosus Variable Childs River Sage Lot Pond Nobska Childs River Sage Lot Pond Nobska Dates 29 August20 September 2013 2 August14 August 2013 Salinity 24.57±0.89 27.04±0.45 31.04±0.05 25.85±0.40 26.33±1.28 31.10±0.32 Nutrient concentrations (µM) NO 3- + NO 25.98±2.58 2.08±0.29 1.85±0.14* 1.07±0.13 1.28±0.15 2.03±0.18** NH 4+ 5.12±1.46 3.12±0.65 1.15±0.09** 2.19±0.01 0.85±0.13 0.57±0.04*** PO 431.70±0.51 1.06±0.12 1.25±0.12 1.55±0.24 0.75±0.15 1.23±0.09** DIN:PO 437.02±2.38 4.99±0.79 2.20±0.3 1.11±0.33 2.29±0.78 2.39±0.55* Table 5.1. Sampling dates, and mean (±se) values of salinity, nutrient concentrations (µM) and DIN:PO4 3ratio during the N isotope equilibration experiments with A. nodosum and F. vesiculosus exposed to water from Childs River, Sage Lot Pond and Nobska (Fig. 5.1). Significant differences among nutrient concentrations in the different sites are shown (*: P≤0.001, **: P≤0.01, ***: P≤0.05). Growth δ15N C:N Species Factor SS df MS F P SS df MS F P SS df MS F P A. nodosum Intercept 91.6 1 91.6 102.2 <0.001 362.8 1 362.8 2282.2 <0.001 7645.8 1 7645.8 77.6 <0.001 Site 0.4 2 0.2 0.2 0.82 0.6 2 0.3 2.0 0.147 351.7 2 175.8 1.8 .181 Time 75.8 4 18.9 21.1 <0.001 3.4 1 3.4 21.4 <0.001 493.5 1 493.5 5.0 .031 Site x Time 70.9 8 8.9 9.9 <0.001 0.6 2 0.3 2.0 0.145 310.1 2 155.0 1.6 .220 Error 26.9 30 0.9 6.2 39 0.2 3840.9 39 98.5 Total 265.5 45 2370.7 45 68363.6 45 F. vesiculosus Intercept 0.3 1 0.3 4.4 0.04 436.9 1 436.9 3031.0 <0.001 3168.8 1 3168.8 145.7 <0.001 Site 0.1 2 0.0 0.6 0.57 0.3 2 0.2 1.1 0.349 26.6 2 13.3 0.6 0.55 Time 4.5 1 4.5 65.8 <0.001 0.2 1 0.2 1.7 0.198 531.0 1 531.0 24.4 <0.001 Site x Time 0.4 2 0.2 2.8 0.08 0.8 2 0.4 2.9 0.068 145.7 2 72.9 3.3 0.049 Error 2.0 30 0.1 4.3 30 0.1 652.5 30 21.7 Total 9.2 36 2742.9 36 37431.6 36 Table 5.2. Results of the general linear model (GLM) univariate procedure to analyze the variability in growth (% d-1), δ15N (‰) or C:N in A. nodosum and F. vesiculosus when grouped by sites (Childs River, Sage Lot Pond or Nobska) and sampling times, as fixed factors. In all cases there was positive growth at the end of the experiment, but maximum growth was recorded after 6 d for A. nodosum and after 12 d for F. vesiculosus . The increase in growth rates was almost continuous during the experiment with F. vesiculosus while growth slowed down between 6 and 12 d of the experiment in the case of A. nodosum . Nevertheless, growth rates of A. nodosum cultured under SLP water were higher than those measured under water from the other sites, while for F. vesiculosus maximum growth rates were observed for CR water (Table 5.2). 105 Chapter 5: Experimental assessment of the macroalgal δ15N study (Table 5.4), and consequently long N turnover times in these macroalgae. Strong isotope fractionation during uptake is not likely to occur. Previous studies with Fucaceae (García-Sanz 2009) and other macroalgae (Cohen and Fong 2005) did not find significant N isotope fractionation related to nutrient concentrations, in contrast with diatoms (Wada and Hattori 1978, Pennock et al. 1996). The rates of change in δ15N in our experiments would have been faster than observed if fractionation were a significant factor, as the light isotopes would have been preferred. For instance, the assayed F. vesiculosus with mean initial δ15N=8.5‰ would have converged to values typical of individuals native of the water origin locations (5.0 to 6.9‰) but they did not show significant changes in their isotopic composition after 12 d. The concentration of ambient N may have also affected changes in macroalgal δ15N. The water employed in the experiments had nutrient concentrations typical of summer in the study area, when uptake by primary producers depletes nutrients (Tomasky et al. 1999). However, N sources, rather than total N concentration determines δ15N in the water and ultimately in primary and secondary producers (McClelland and Valiela 1998b, Viana and Bode 2013). Experiments with other species showed that macroalgal δ15N did not change with water N concentrations as long as the δ15N of dissolved N was constant (Cohen and Fong 2005, GarcíaSanz 2009). Furthermore, nutrient uptake in F. vesiculosus is less dependent on substrate concentration than in green or red algae (Pedersen and Borum 1997). In our experiment with water of different origins, the low concentrations of dissolved N did not prevent the individuals of both species from growing in weight and maintaining C:N values characteristic of non N-limited algae (Niell 1976), thus suggesting that the slight changes in δ15N were not a direct consequence of water N concentration. The relatively high nitrogen content (1.2±0.3% for A. nodosum , 1.4±0.1% for F. vesiculosus ) and the enriched δ15N values of macroalgae at the starting point could have also influenced isotopic equilibration. Slow-growing brown macroalgae usually rely on their internal N pools during periods of low nutrient supply, as in summer seasons in temperate areas (Lehvo et al. 2001, Villares et al. 2013). During these periods growth rates and external nutrient demand are lowered while the macroalgae, eventually profiting from high light levels, develop carbon reserves, thus increasing tissue C:N, as observed in our experiments (Fig. 5.2). Naldi and Wheeler (2002) also observed that high total N content of thalli influenced nitrate uptake rates in green 112 and red macroalgal species. Low external N demand along with large difference in δ15N values between the macroalgal tissue and the surrounding water (as suggested by the δ15N values of native macroalgae), may be the main determinants of the rate of isotopic equilibration in our incubations with F. vesiculosus . Other experiments with transplanted individuals of this species in the field also found small or no changes in their tissue δ15N after days of incubation (Deutsch and Voss 2006). In contrast, and despite the longer turnover time, A. nodosum started to show differences in δ15N after 12 days of incubation, likely because the initial values for this species were much lower than those for F. vesiculosus . N uptake and turnover along the thallus The results of the enrichment experiments showed that both species do not transport recently absorbed N along their thallus, at least during 24 h after uptake (Fig. 5.4). Despite their internal structure (i.e. symplastic pathway) suited for transport (Raven 2003), only carbon photosynthetic assimilates were reported to translocate along the thallus of some Fucaceae (Diouris and Floc´h 1984). Inorganic nitrogen transport, however, was reported for other brown macroalgae, such as Laminariales (Mizuta et al. 1996, Hepburn et al. 2012). These algae have nutrient requirements different from those of Fucales as they show basal meristematic growth, which means that they grow where the blade and the stipe meet (Lobban and Harrison 1994). In contrast, Fucales show mostly apical growth and therefore concentrate N demands in the tips of the thallus (Topinka Bigelow 1978), although as demonstrated by our enrichment experiment (Fig. 5.4c), all sections of the thallus are able to take up inorganic N from the water. As N transport have relatively high energy and oxygen requirements (Raven 2003), this process can be avoided if both assimilation and uptake occur in the same part of the thallus. In Laminariales, N uptake and assimilation occur at different rates in the different parts of the thallus, deriving in gradients along the frond (Mizuta et al. 1996). Despite their apical growth, variation in δ15N values along the thallus has been reported for Fucus species (Savage and Elmgren 2004, Raimonet et al. 2013) and in the present study (Fig. 5.3). If transport is excluded, such intraindividual variation might be due to differential uptake and growth, or to isotope fractionation in the different sections of the thallus. 113 Chapter 5: Experimental assessment of the macroalgal δ15N In the enrichment experiment we showed that both species were able to incorporate dissolved nitrogen when submerged (Fig. 5.4). The process of nitrogen uptake and assimilation in macroalgae involves transport from the water column and then assimilation into organic compounds, followed by incorporation into proteins and macromolecules for growth (McGlathery et al. 1996). Growth is the most important N sink in macroalgae. In mature segments, N demand for structural pools is not as important as in growing tips, this would explain why N uptake at the non-growing segments was only half the uptake rate measured at the tips of F. vesiculosus when all the frond was submerged (Table 5.4). For A. nodosum there was also a marked difference in the uptake rates of the tip and those of the mature segments, at least when only one of the sections was submerged. These results agree with studies reporting higher N uptake in apical fronds and whole young plants or germlings and lowest in slower-growing older fronds and stipes of F. spiralis (Topinka Bigelow 1978, Rosenberg et al. 1984) and differential 15N enrichment along thalli regions of F. vesiculosus (Döhler et al. 1995). Non-apical segments of A. nodosum and F. vesiculosus individuals can store N to use in metabolic processes other than growth. For instance, N can be accumulated as inorganic (NO3 - and NH4 +) and organic compounds (as phycobiliproteins) and can be found in algal pigments (Hanisak 1983) although NH4 + storage capacity is limited due to toxicity (Haines and Wheeler 1978, Lotze and Schramm 2000). The net short-term N uptake recorded along the thallus implies that δ15N values of different sections would change with the isotopic composition of the surrounding water at rates depending on their initial δ15N value, and of the processes affecting isotope fractionation within each section. Nitrogen release, both in organic and inorganic forms, has been observed for some green and red macroalgae (Naldi and Wheeler 2002, Tyler and McGlathery 2006) and was interpreted as the result of isotopic equilibration of internal and external pools (Fujita et al. 1988) or to stress due to sudden changes in the proportion of different N sources (Naldi and Wheeler 2002). Even when fractionation during uptake, resulting in tissue δ15N values lower than those of the water, is not likely (García-Sanz 2009), the release of preferentially light N isotopes may explain the higher enrichment of the tip sections compared to other parts of the thallus, as found in our experiments (Figs. 5.3 and 5.4) and in other studies (Raimonet et al. 2013). As far as we know, there are no reports of N release in the species considered in our study, but it can be expected that this process is restricted to the most metabolically active tissues. 114 Implications for the use of A. nodosum and F. vesiculosus to monitor landderived nitrogen sources The results of the present study are of application when using A. nodosum and F. vesiculosus to study the impact of anthropogenic N sources on littoral ecosystems both analyzing native populations and in incubation experiments, the latter applicable when these species are not naturally present in the impacted area. Taking advantage of the apical growth and long life span of both species, Savage and Elmgren (2004) interpreted δ15N values in different sections of the thallus of F. vesiculosus in a retrospective study to monitor changing N loadings. The underlying assumptions were that annual growth occurred only at the tips and, by knowing the rate of growth, each section of the thallus could be dated and associated to a particular period of exposure to the ambient N. Thus, δ15N of the sections would reflect past N sources if mature segments do not equilibrate N contents with the surrounding water and if there is no transport of N along the thallus. Other studies, however, questioned this application for retrospective studies as they found contrasting patterns of change along the thallus that could not be related to ambient N (Raimonet et al. 2013, Carballeira et al. 2014). The enrichment experiment in this study demonstrated that all sections of the thallus of both species take up N from the ambient water when submerged. Even when there was no transport of the N along the thallus and the rates of uptake at the mature parts of the frond were lower that at sections located at or near the tip this uptake would affect the δ15N of the sections. These results explain why previous studies found contrasting patterns of change of δ15N along the thallus of F. vesiculosus (Carballeira et al. 2014) as the δ15N of each section changes with the isotopic composition of the water at different rates. Therefore, it is not possible to obtain unbiased estimates of past N sources from the δ15N of different sections of the thallus of these macroalgae. Furthermore, determinations of δ15N from pooled samples of different sections would produce δ15N values resulting from a mixture of past and present N sources, depending on the amount of matter from sections with different turnover rates. Pooled samples of the whole individual can be also be misinterpreted if individuals of different lengths (i.e. ages) are used. However, δ15N of the tips can be used as monitors of N sources in the ambient water averaged over scales of 15 days ( F. vesiculosus ) and up to 6 months ( A. nodosum ). This range of integration times is particularly appropriate to differentiate chronic pollution from point discharges that may have little impact on the macroalgae. 115 Chapter 5: Experimental assessment of the macroalgal δ15N Besides the use of natural populations, these macroalgae can be used in transplantation or laboratory experimental incubations with different water types to determine potential impacts of different N sources (Deutsch and Voss 2006). In this case, the turnover and equilibration times of the tips, as determined in the present study, need to be taken into account when determining the duration of the incubations. Otherwise the results will not reflect the actual impact of the ambient N sources. 116 119 * Viana IG, Bode A (in review) Variability in δ15N of intertidal brown algae along a salinity gradient: differential impact of nitrogen sources. Sci Total Environ Variability in δ15N of intertidal brown algae along a salinity gradient: differential impact of nitrogen sources* Abstract While it is generally agreed that δ15N of brown macroalgae can discriminate between anthropogenic and natural sources of nitrogen, this study provides new insights on net fractionation processes occurring in some of these species. The contribution of continental and marine sources of nitrogen to benthic macroalgae in the estuaryria system of A Coruña (NW Spain) was investigated by analyzing the temporal (at a monthly and annual basis) and spatial (up to 10 km) variability of δ15N in the macroalgae Ascophyllum nodosum and three species of the genus Fucus ( F. serratus , F. spiralis and F. vesiculosus ). Total nitrate and ammonium concentrations and δ15NDIN, along with salinity and temperature in seawater were also studied to address the sources of such variability. Macroalgal δ15N and nutrient concentrations decreased from estuarine to marine waters, suggesting larger dominance of anthropogenic nitrogen sources in the estuary. However, δ15N values of macroalgae were generally higher than those of ambient nitrogen at all temporal and spatial scales considered. This suggests that the isotopic composition of these macroalgae is strongly affected by fractionation during uptake, assimilation or release of nitrogen. Besides, the absence of correlation between macroalgal and water samples, suggests that the δ15N of the species considered can be used to determine the impact of the different nitrogen sources integrated over long-time periods, but not for monitoring shortterm changes. keywords: Fucus Ascophyllum monitoring DIN coast salinity gradient Chapter 6: Variability in δ15N of intertidal brown algae Introduction Coastal areas integrate the nitrogen inputs from continental and marine sources. The growing anthropogenic pressure in these areas has increased dissolved inorganic nitrogen (DIN) concentrations in continental sources in comparison with seawater. Nitrogen concentrations could also be affected by other nitrogen inputs, as atmospheric deposition and upwelled seawater, and by processes as nitrogen uptake by primary producers, regeneration and movement of dissolved organic nitrogen (DON) at the sediment surface (Fry 2002). Therefore, monitoring nutrient concentrations is not enough to identify the origin of DIN in coastal areas (Sebilo et al. 2006). To overcome this constrain, the ratio of nitrogen stable isotopes (δ15N) has been increasingly used as marker of anthropogenic nutrient loading in the last years (McClelland and Valiela 1998b, Ahad et al. 2006, Deutsch and Voss 2006, Schubert et al. 2013). The direct measurement of stable isotopes on DIN has been widely used and characteristic isotopic signatures were identified for different nitrogen origins (Heaton 1986, Ahad et al. 2006, Raimonet et al. 2013, Viana and Bode 2013). Human and animal wastewaters are enriched in 15N relative to seawater because of strong isotopic fractionation during nitrification and volatilization in the case of NH4 +, or denitrification in the case of NO3 − (Mariotti et al. 1981). In contrast, synthetic fertilizers are depleted in 15N due to the atmospheric origin of the fixed nitrogen (Heaton 1986). Consequently, δ15N values of marine organisms show large variability in human impacted sites (McClelland and Valiela 1998b, Fry et al. 2003, Bode et al. 2014). Macroalgae have been traditionally employed as biomonitors of eutrophication because their growth is rapidly enhanced by nutrient inputs (McClelland and Valiela 1998b, Piñón-Gimate et al. 2009). The δ15N of the different macroalgal species shows a large variation, from 0.2 to 50.1‰ (Dailer et al. 2010). Besides the generalized use of macroalgae for monitoring nitrogen sources, only recent studies have addressed the influence of intrinsic or external factors on the variability of the isotopic values. The main intrinsic factors affecting variability of macroalgal δ15N are the preferential mobilization of light isotopes (isotopic fractionation) during uptake, excretion and metabolic reactions (Teichberg et al. 2008) and intra-frond variability (Savage and Elmgren 2004, Raimonet et al. 2013, Carballeira et al. 2014, Viana et al. in review a). External factors are littoral position (Kim et al. 2013) or light (Dudley et al. 2010). 121 Chapter 6: Variability in δ15N of intertidal brown algae 128 Figure 6.2. Variation in (a) temperature (t, °C), (b) salinity (S), (c) DIN (total nitrate and ammonium, μM), (d) δ15N-DIN (total nitrate and ammonium, ‰) and (e) δ15N in macroalgae (mean±se, F. vesiculosus and A. nodosum , ‰) at Ría do Burgo and Mera (open squares, Fig. 6.1) from October 2010 until January 2012. 129 Differences in the isotopic composition of A. nodosum and F. vesiculosus among different years (interannual variability) were studied by Kruskall-Wallis (K-W) tests and a posteriori Dunnett-C tests for multiple comparisons. All the above mentioned tests were performed with SPSS Statistical Software. Results Seasonal variability The main changes observed through the year (Fig. 6.2) were related with decreases in salinity during winter due to the higher volume flow of the river Mero, which affected the site at Ría do Burgo, but had negligible influence at Mera (Fig. 6.2b). Despite the variability observed during these winter months, similar and low concentrations of total nitrate were observed at both sites during the rest of the year. The inverse correlation between total nitrate concentrations and temperature or salinity indicates a minor influence of upwelling compared to river inputs at both sites (Table 6.1). In contrast, ammonium concentrations were uncorrelated with other variables and were lower than those of total nitrate from November 2010 to June 2011, while in summer they were higher and more variable (Fig. 6.2c). The isotopic composition of ammonium and total nitrate were positively correlated (Table 6.1). The highest δ15N values were observed in summer and the lowest during winter, particularly at the oceanic influenced site of Mera (Fig. 6.2d). Despite the similar pattern, total nitrate (but not ammonium) δ15N was positively correlated with temperature (Table 6.1). Considering all the observations there were not significant differences between sites for any of the water variables considered (M-W test, P>0.05, n=26, and n=23 for isotopic values in DIN). Chapter 6: Variability in δ15N of intertidal brown algae tSTNammonium δ15NTNδ15Nammonium %NA. nodosum %NF.vesiculosus δ15NA. nodosum δ15NF.vesiculosus t-0.026 0.000 0.0920.038 0.224 0.006 0.000 0.112 0.350 S0.436 -0.034 0.434 0.154 0.191 0.8790.155 0.121 0.033 TN-0.670 -0.417 -0.612 0.781 0.920 0.007 0.000 0.070 0.286 ammonium 0.337 0.160-0.104 -0.596 0.744 0.175 0.215 0.572 0.489 δ15NTN0.435 0.307 -0.061 0.117 -0.000 0.5760.305 0.5460.268 δ15Nammonium 0.264 0.283 -0.022 0.072 0.840 -0.286 0.511 0.0580.149 %NA. nodosum -0.741 0.0490.732 -0.420 -0.217 -0.400 -0.000 0.3060.986 %NF. vesiculosus -0.758 -0.300 0.785 -0.263 -0.235-0.152 0.912 -0.150 0.034 δ15NA. nodosum -0.483 -0.472 0.539 -0.182-0.233 -0.650 0.3080.423 - 0.255 δ15NF.vesiculosus -0.200 -0.436 0.227 0.148 -0.253 -0.3260.005 0.418 0.341 - Table 6.1. Spearman ρ correlation coefficients (lower semimatrix) and significance (upper semimatrix) between environmental and isotope composition variables measured at the intertidal stations of Ría do Burgo and Mera (from October 2010 until January 2012). Significant values (P<0.05) appear in boldface. TN: total nitrate, %N: mass percent content in nitrogen of the macroalgal samples, t: temperature, S: salinity. 130 -8 -4 0 4 8 O N D J F M A M J J A S O N D 15Nalga-15NDIN (‰) A. nodosum Ría do Burgo -8 -4 0 4 8 O N D J F M A M J J A S O N D 15Nalga-15NDIN (‰) total nitrate ammonium F. vesiculosus Ría do Burgo -8 -4 0 4 8 O N D J F M A M J J A S O N D 15Nalga-15NDIN (‰) F. vesiculosus Mera 2010 2011 a b c Figure 6.3. Monthly variation (from November 2010 to November 2011) of the difference between the mean δ15N (‰) of Fucus spp. and total nitrate and ammonium at Ría do Burgo (a) and Mera (c) and the difference between the mean δ15N (‰) of A. nodosum and total nitrate and ammonium at Ría do Burgo (b). 131 The seasonal δ15N variability observed in macroalgae was lower than in DIN (Fig. 6.2e). Both macroalgal species showed low δ15N values during spring and summer (minimum in July) and high values in fall and winter (maximum in December 2011). Besides the similar variation along the year, δ15N of F. vesiculosus and A. nodosum were not correlated (Table 6.1). On the contrary to what was observed with water samples, δ15N of F. vesiculosus showed higher mean values at Ría do Burgo than at Mera (M-W test, P<0.001, n=26), although the variability at the latter was higher. This variable was negatively correlated with salinity and positively with the nitrogen content. In contrast, δ15N values of A. nodosum were not correlated with any other variable (Table 6.1). In Ría do Burgo, mean δ15N values of F. vesiculosus exceeded those of both total nitrate and ammonium for most of the year, except for ammonium in July and September 2011 (Fig. 6.3a). A. nodosum followed a similar variation pattern but in this species the δ15N values were lower than ammonium and total nitrate also in November 2011 (Fig. 6.3b). In contrast, mean δ15N of F. vesiculosus in Mera was only slightly above δ15N of total nitrate and ammonium between January and June 2011, and also in November 2011, while values in November and December 2010 and between July and October 2011 where lower than those of DIN (Fig. 6.3c). Interannual variability Stable nitrogen composition of F. vesiculosus and A. nodosum showed different interannual variability patterns at both sampling sites (Fig. 6.4). While δ15N of F. vesiculosus did not vary significantly between years at Ría do Burgo, there was a significant increase in 2013 compared to values recorded in 2006 and 2011 at Mera (K-W and Dunnett-C tests, Fig. 6.4a). In turn, A. nodosum δ15N decreased from 2010 to recent years (K-W and Dunnett-C tests, Fig. 6.4b). Spatial variability The estuary-ria system of A Coruña showed a marked spatial gradient in salinity but not in temperature (Fig. 6.5a, b). The range of variation of salinity (9 to 34) indicates the differential influence of terrestrial and oceanic waters along the estuarine mixing zone. Annual averaged concentrations of total nitrate (but not ammonium) followed an inverse relationship with salinity, with the highest values near the river end and the lowest values in the outer bay (Fig. 6.5c). Therefore, considering individual paired samples, only total nitrate concentrations were significantly correlated with salinity Chapter 6: Variability in δ15N of intertidal brown algae 132 (Spearman ρ=-0.684, P<0.001, n=68) indicating the conservative mixing of this nitrogen source between riverine and marine waters along the transect considered. In contrast, δ15N of both DIN forms showed a small decrease along the estuary while their corresponding average values were almost constant in the bay (Fig. 6.5d). Overall, δ15N of ammonium was higher than δ15N of total nitrate (M-W test, P<0.05, n=68) and both isotopic signatures were significantly correlated (Spearman ρ=0.935, P<0.001, n=68). 6 7 8 9 10 2006 2007 2008 2009 2010 2011 2012 2013 δ15N (‰) Ría do Burgo Mera F. vesiculosus * 6 7 8 9 10 2006 2007 2008 2009 2010 2011 2012 2013 δ15N (‰) A. nodosum * b a Figure 6.4. Interannual variation of δ15N (mean±se, ‰) values of F. vesiculosus at Ría do Burgo and Mera (a) and A. nodosum at Ría do Burgo (b). *: significant differences (Kruskal-Wallis and Dunnett-C tests, *: P<0.05). 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 δ15N-DIN (‰) total nitrate ammonium d 0 20 40 60 80 100 120 140 0 2 4 6 8 10 12 DIN (μM) total nitrate ammonium c 10 15 20 25 0 2 4 6 8 10 12 t (ºC) a 0 10 20 30 40 0 2 4 6 8 10 12 S b Relative distance (km) Figure 6.5. Variation of the mean values (±se, from July 2009 to January 2012, n=14) in (a) temperature (t, °C), (b) salinity (S), (c) DIN (total nitrate and ammonium, μM) and (d) δ15N-DIN (total nitrate and ammonium, ‰) along the relative distance (km) from the reference point (Fig. 6.1). Relative distances greater than 5 km (salinity >34) correspond to the bay, while relative distances <5 km correspond to the estuary. 133 Chapter 6: Variability in δ15N of intertidal brown algae 134 Macroalgal δ15N followed a spatial pattern similar to δ15N in DIN, with values decreasing from the inner estuary to the outer bay (Fig. 6.6). Macroalgae from the estuary had nitrogen isotopic values closer to the δ15N of an effluent of a water treatment plant. While isotopic values within the bay decreased towards the isotopic values of deep oceanic water from the mouth of the Bay of A Coruña (Fig. 6.6). δ15N values of Fucus spp. from the estuary were significantly higher than those from the bay 0 6 12 18 0 2 4 6 8 10 12 δ15Nalga (‰) left right Fucus spp. a 0 6 12 18 0 2 4 6 8 10 12 δ15Nalga (‰) A. nodosum b Relative distance (km) Figure 6.6. Variation of the δ15N values (mean±se, ‰,) of Fucus spp. (a) and A. nodosum (b) along the relative distance (km) from the reference point (Fig. 6.1) at the left and right banks in July 2013. Relative distances greater than 5 km correspond to the bay, while relative distances <5 km correspond to the estuary. δ15N values of end members in the area are represented, the dark gray bar shows the δ15N values of total nitrate (16.6±1‰, n=3) of a water treatment plant effluent, while the light gray bar shows the δ15N of total nitrate (6.02±0.28‰, n=24) and ammonium (5.73±0.28‰, n=24) of deep oceanic water (70 m) from the mouth of the Bay of A Coruña. -3 -2 -1 0 1 2 3 0-0.5 1-1.5 2-2.5 3-3.5 4-4.5 5-10 15Nalga-15NDIN (‰) relative distance interval (km) total nitrate ammonium Fucus spp. a -3 -2 -1 0 1 2 3 0-0.5 1-1.5 2-2.5 3-3.5 4-4.5 5-10 15Nalga-15NDIN (‰) relative distance interval (km) A. nodosum b Figure 6.7. Variation of the difference between the mean δ15N (‰) in total nitrate and ammonium and Fucus spp. (a) and A. nodosum (b) at increasing relative distance intervals (0.5-km bins) from the reference point (Fig. 6.1). The relative distance interval greater than 5 km correspond to the bay, while distance intervals <5 km correspond to the estuary. 135 (M-W test, P<0.001, n=88). In the latter zone, there was also a significant difference between the δ15N values of Fucus spp. collected at the different banks, with specimens from the right bank more enriched than those from the left bank (M-W test, P<0.001, n=33). However, there were no significant differences between banks in the isotopic composition of specimens of Fucus spp. or A. nodosum inside the estuary (M-W tests, P>0.05, n=55 and n=29 for Fucus spp. and A. nodosum , respectively). Chapter 6: Variability in δ15N of intertidal brown algae 136 In general, macroalgae were enriched in 15N relative to DIN along the study area, with the exception of the depleted values measured for Fucus spp. in the outer bay (Fig. 6.7a). The enrichment increased from the river end (0 km in the reference distance scale) to the center of the estuary (2.5 km) and decreased thereafter towards the outer bay (up to 10 km). This enrichment was higher for Fucus spp. than for A. nodosum , the latter showing even 15N depletion relative to ammonium at some sites (Fig. 6.7b). When compared in these 0.5-km intervals, δ15N of Fucus spp. was significantly correlated with isotopic values of ammonium or total nitrate (Spearman ρ=0.864, P<0.001, n=11, for both total nitrate and ammonium). In contrast, no significant correlation was found between isotopic values of DIN and A. nodosum because this macroalgae is restricted to the estuarine site with less isotopic variability in algae and DIN. Discussion Correspondence between δ15N in algae and water The isotopic values of native fucoid species were not correlated with concurrent isotopic values of DIN (Table 6.1). This result has been also reported in other studies and was attributed to isotopic fractionation processes (Deutsch and Voss 2006, Raimonet et al. 2013) or to the rapid mixing of different nitrogen sources in estuaries (Derse et al. 2007). Fractionation can be observed in almost all physical processes and chemical reactions involved in the cycle of nitrogen (Mariotti et al. 1981). On the other hand, water mixing is an effective mechanism to dilute isotopically enriched DIN discharged to coastal waters (Ahad et al. 2006, Raimonet et al. 2013). Periodic mixing events (as in most tidal regimes) would enhance high frequency variability in the isotopic composition of DIN that may not be tracked by the macroalgae. The net isotopic fractionation of macroalgal N can be defined as the result of all processes involved in the incorporation of nitrogen to macroalgal tissues, i.e. uptake from the water, assimilation into organic compounds, storage and release of dissolved forms. Fractionation during uptake leads to depletion of δ15N in the algal tissues compared to DIN (Pennock et al. 1996). However, there is some controversy about the importance of this type of fractionation for macroalgae, as there are examples of no fractionation (Cohen and Fong 2005, García-Sanz 2009) or significant fractionation for some compounds (Kaldy 2011). In our study, fractionation during uptake may have been limited to the few cases with macroalgal δ15N lower than δ15N 137 in DIN sources, as observed for Fucus spp. in the bay, both during late summer and fall (Fig. 6.3) and for the spatial distribution study in July 2013 (Fig. 6.7). Apart from these cases, the lack of significant correlations between the δ15N of macroalgae and DIN concentrations suggests that uptake fractionation was not the main mechanism explaining δ15N in the studied macroalgae. Fractionation can also result from the release of nitrogen from the macroalgae. In this case the macroalgae would be more enriched than DIN in the surrounding water, as they would preferentially release the light isotope (Wada et al. 1975). Release of both inorganic and organic compounds has been detected in several macroalgal species (Fong et al. 2004, Tyler and McGlathery 2006). It was attributed to a response to different stressors (Young et al. 2009) and particularly to emersion (Kim et al. 2013). As far as we know, there are no reports of nitrogen release in the species considered in this study, but our results suggest that this process is enhanced in the zone with large changes in salinity. The frequent changes in salinity experienced by the central part of the estuary (error bars, Fig. 6.5) would act as an stressor agent. Nitrogen content of macroalgae is mainly related with total nitrate concentrations (Table 6.1), as this nutrient is the preferred DIN source for growth (Pedersen and Borum 1997). Therefore the changing relation between macroalgal and DIN isotopic values is not expected to be related with the preference for other dissolved nitrogen forms, as DON or NH4 +. Even though ammonium was preferred, the similar isotopic values of both DIN components would not affect the δ15N values in macroalgae. Besides, there was not DIN depletion in the area, so the preferred use of other dissolved forms might not be expected (Tyler et al. 2005). The variability in the isotopic enrichment of macroalgae relative to DIN found in this study implies that, at any given time, both compartments are in different states of isotopic exchange. These species have slow apical growth rates, varying from 2.2 to 1.4 cm mo-1 in F. vesiculosus (Viana et al. in review b) to 1.6 cm mo-1 in A. nodosum (Viana et al. in press). Therefore, the 1-cm sample used in this study corresponds to a 14-day and 21-day exposure time of F. vesiculosus and A. nodosum respectively (Viana et al. in press, in review b). Besides, this would imply the correspondence of isotopic values in macroalgae with isotopic values of DIN of previous months, but this was never observed in this study (Fig. 6.2d, e). Therefore, correspondence between δ15N values of these macroalgae and DIN cannot be expected because the Chapter 6: Variability in δ15N of intertidal brown algae 144 In the study area, macroalgae are enriched in heavy isotopes at sites of large influence of upwelling while they are depleted where the upwelling has lower influence. Latitudinal variability was also observed in other macrophytes (Christiaen et al. 2013), and it was related with upwelling processes in some biota in South Africa (Hill and McQuaid 2008). This variability can be predicted to some extent thus facilitating the comparison of isotopic signatures of organisms collected in regions of changing N inputs. Besides the natural gradient, the anthropogenic influence also accounted for part of the variance observed in δ15N for both species, with an increase of δ15N with the size of the urban population. Previous studies in Galicia have shown that enriched macroalgal δ15N values could be related to local anthropogenic N inputs (Bode et al. 2011b) particularly inside the rias (Viana et al. 2011), but no quantitative estimations of the variability in N sources were made in this or other areas with these species. Nevertheless, isotopic signatures in macroalgae are not simply related to the size of the human population, as for instance, the variability in small populations is very high while mean values observed near large cities (>15,000 inhabitants) are relatively less variable. The reasons for isotopic variability observed in macroalgal samples close to small urban nuclei is difficult to explain due to the absence of a direct relationship between δ15N of concurrent macroalgal and water samples. Both external and internal factors could explain these differences, as the different wastewater treatments applied, their efficiency, or the influence of macroalgal metabolism. Therefore, from this content block some recommendations can be extracted when establishing monitoring protocols or when comparing isotopic values at large regional or latitudinal scales, from the same or from different studies. First, a good knowledge of local or regional natural factors affecting isotopic signatures is needed for the interpretation of the results, as the existence of upwelling processes, or high N2 fixation in the area (as in Lamb et al. 2012). Second, it is also necessary to consider that enriched isotopic values are not simply related with higher anthropogenic pressure. 145 Applying the use of N isotopic values in macroalgae to retrospective studies: a solution for long-term monitoring programs? Long-term monitoring is needed to track the ecological status of ecosystems in time (Koslow and Couture 2013) or to contextualize current observations. However, obtaining reliable and long-time series generally requires a careful sampling plan implemented during decades at sites sensitive to anthropogenic influence. Consequently there are only a few examples of time series using stable isotopes and only for selected sites (e.g. Viana et al. 2011). Therefore, there were great expectations on using the old-growth parts of these macroalgae for retrospective studies, which would help to interpret the impact of anthropogenic derived N (Savage and Elmgren 2004, Raimonet et al. 2013, Carballeira et al. 2014) or other contaminants (Stengel et al. 2005, Heldal and Sjøtun 2010). Moreover, this approach would allow reducing the sampling effort in monitoring programs as it would allow reducing sampling frequency (Carballeira et al. 2014). However, when using these species as retrospective biomonitors two main assumptions remained untested until now: i) to feasibly relate macroalgal segments with different exposure periods (months, years), and ii) to verify if the macroalgal physiology could alter the isotopic signature of non-growing segments. To relate one macroalgal segment with a particular exposure period, two approaches can be done with the selected species. First, the number of bifurcations in F. vesiculosus or the number of gas bladders in A. nodosum could be used for dating different growing periods (Savage and Elmgren 2004, Stengel et al. 2005). These markers of age produced different results. In contrast to previous studies (Savage and Elmgren 2004) we could not obtain a clear relationship between the number of bifurcations in F. vesiculosus and the length or age of the individuals (Fig. 7.1). The number of bifurcations for this species varies mainly with exposure and salinity rather than with age (Jordan and Vadas 1972, Kalvas and Kautsky 1993, López-Rodríguez et al. 1999). Therefore the level of branching of this species cannot be used as a proxy for age and therefore for retrospective studies of past nitrogen sources using stable isotopes. On the contrary, our study confirmed the annual appearance of gas bladders in A. nodosum after the first year of life (Niell 1979). Because of this feature, the gas bladder of this species could be used for dating yearly segments of their thallus. Chapter 7: General discussion 146 On the other side, the growth curves of the species can be estimated. This would be particularly important for F. vesiculosus due to the absence of a clear relationship between the number of dichotomies and age. The growth curves for different F. vesiculosus and A. nodosum populations obtained in Chapters 3 and 4 were intended for application to retrospective studies. With these curves we are able to estimate the time required for individuals from these populations for reaching a specific length, and these estimations can be extended to populations living in similar environments (i.e. estuarine or semi-exposed sites). Because these species have apical growth, if the isotopic composition of the growing segment remains unchanged after growth we could estimate the N sources at the time of growth by analysing different segments of the thallus. For the second assumption, the enrichment experiments described in Chapter 5 demonstrated that all sections of the thallus of both species take up N from the ambient water when submerged. Even when there was no transport of N along the thallus, the results show that this uptake would affect the δ15N of non-growing sections. The differential values along the thallus observed in previous studies cannot be a direct consequence of the exposure to different δ15N-DIN values during previous years. Therefore, it is not possible to obtain feasible estimates of past N sources from the δ15N of different sections of the fronds of A. nodosum or F. vesiculosus . Even though long-term monitoring programs cannot be based on the retrospective analysis of individuals, the growing tips can still be used for the establishment of these programs (Viana et al. 2011). Due to the different values observed along the thallus, it is important to limit the use to the apical tip and also specify the length of the segment, as this may influence the exposure period and hence the interpretation of the results. From the growth curves estimated in Chapters 3 and 4, the reflected period by the apical tips with different lengths can be calculated (Table 7.1). Until now, most studies using these species for monitoring purposes have considered the use of the growing tips, although different length segments have been used (Table 1.1). Moreover, the estimated growth rates are also age-dependent in these species (Chapters 3 and 4). Therefore the total length of the individuals used should be similar, in order to guarantee that the segments considered are reflecting the same growing periods. Individuals of 10 and 60 cm long for F. vesiculosus and A. nodosum respectively 147 Chapter 7: General discussion would be adequate as at that time they show exponential growth rates and they have not reached the length when the probability of suffering breakages is higher. 0 5 10 15 0.0 0.5 1.0 1.5 2.0 Dichotomies Cohort 1 0 5 10 15 0.0 0.5 1.0 1.5 2.0 Dichotomies Cohort 2 0 5 10 15 0.0 0.5 1.0 1.5 2.0 Dichotomies Cohort 3 0 5 10 15 0.0 0.5 1.0 1.5 2.0 Dichotomies Cohort 1 0 5 10 15 0.0 0.5 1.0 1.5 2.0 Dichotomies Cohort 2 0 5 10 15 0.0 0.5 1.0 1.5 2.0 Dichotomies Cohort 3 Age (years) Ría do Burgo Mera Figure 7.1. Mean (dots) and range (shaded area) of the number of dichotomies in the thallus observed for the three cohorts of F. vesiculosus at Ría do Burgo and Mera (Chapter 3). 148 Applying the use of δ15N in macroalgae for the study of the local influence of N sources. The case study of the Ría de A Coruña Differential δ15N values in macroalgae in monitoring studies at local scales have been attributed to the differential impact or mixing of the sources. The laboratory experiments and the case study at the Ría de A Coruña highlighted the importance of external and macroalgal intrinsic factors for the interpretation of macroalgal δ15N values at local scales. Among external factors some clues about where, when and how often to do the samplings can be deduced. The results revealed that spatial variation of δ15N in macroalgae was of paramount importance, even when changes in isotopic values of DIN were not significant at this scale. Both macroalgae and DIN reflected that the estuary (Ría do Burgo) was more impacted by isotopically enriched N than the outer bay of A Coruña. This variability suggested the larger influence of anthropogenic N in the estuary while marine N would dominate in the bay. Therefore, to account for the general status of the ria or estuary a minimum of two sampling sites are needed: one in the inner part, where salinity is <34, and other site where salinity is >35. For selecting the position of these sites, the circulation of the water must also be considered. In contrast, isotopic DIN values varied mainly with time, at least at the seasonal scale examined, while macroalgal δ15N variation was not that high. These differences might be explained as a direct consequence of the integration times necessary to F. vesiculosus A. nodosum 1 cm 14 21 3 cm 42 63 5 cm 71 106 10 cm 177 222 Table 7.1. Growth time (days) reflected by individuals of 10-cm long of F. vesiculosus and 60-cm long of A. nodosum from Ría do Burgo depending on the length of the apical segment considered. The estimated time is based on the growing curves obtained for cohort 1 in F. vesiculosus (Chapter 3) and for A. nodosum at this site in 2010 (Chapter 4). 149 Chapter 7: General discussion acquire the δ15N signature of DIN for both macroalgal species (Chapter 5). When establishing monitoring protocols, this variability might also be considered. Due to the low seasonal variability observed, samplings can be restricted to a few months, or at least one month along the year. Taking into account the growth rates of the species, a 6-month periodicity will be ideal for covering the annual variability at a particular site. But if sampling is done at an annual basis (as in Chapter 6), the same sampling month should be considered. Taking into account the observed variability, sampling surveys should be established between spring and summer, when higher isotopic values are observed in estuarine sites and most oceanic influenced sites respectively. Besides the use of natural populations, these macroalgae can be used in transplantation or laboratory experimental incubations with different water origins to determine potential impacts of different N sources (Chapter 5). Apart from the previous recommendations that might be also considered in these studies, in this case, it is important to consider the nitrogen turnover rates and equilibration times of the tips for these species when establishing the duration of the incubations. The δ15N of the tips of macroalgae integrate N sources in the ambient water over scales of 15 days ( F. vesiculosus ) and up to 6 months ( A. nodosum ). With lower incubation times the results will not reflect the real impact of the ambient N sources. Concerning the intrinsic factors of macroalgae, the absence of a direct and conservative relationship between macroalgal and water δ15N complicates the quantitative interpretation of isolated macroalgal δ15N (as shown in Chapter 2). In Chapter 6 this relationship was shown to be variable in time and space, although a concordance was observed when values were averaged over a large spatial area. The integration times necessary for macroalgae to acquire the δ15N signature of DIN might explain this difference. This is due to the fact that they are not reflecting the same N composition at a particular time. Our results illustrate the need to address the variability of isotopic signatures along a gradient. Many studies using δ15N values in macroalgae to detect the area of influence of point source effluents used gradients that reported decreasing values with distance from the emission source (Costanzo et al. 2001, Savage and Elmgren 2004, García-Sanz et al. 2010, Carballeira et al. 2013). The problem arises when diffuse or unknown sources are present in a large salinity gradient, as shown in Chapter 6. In these cases, repeated studies of δ15N in the macroalgae along the main spatial gradient are required to account for the unknown variability in the N sources. 150 Besides integration times, isotopic fractionation during N metabolism may be more important to determine the observed δ15N composition than temporal or even spatial variability. When estimating the fractional contribution of anthropogenic and natural sources it is generally assumed that these macroalgae do not exhibit isotopic fractionation (e.g. Savage and Elmgren 2004). If the fractionation factor is 0 20 40 60 80 100 -2 0 3 % contribution Isotopic enrichment factor (Ɛ) 0 20 40 60 80 100 -2 0 3 % contribution Isotopic enrichment factor (Ɛ) b a Figure 7.2. Percent contribution of anthropogenic (black bars) and marine nitrate (gray bars) to Fucus spp. (a) and A. nodosum (b) δ15N values at a site at Ría do Burgo. The contribution was estimated from a mixing model with two end-members: δ15Nma= ƒaδ15Na+ (1-ƒa) δ15NmƐ, where the δ15Nma is the isotopic composition of macroalgae, ƒa and 1-ƒa the contributions of anthropogenic and marine sources respectively, and δ15Na and δ15Nm the anthropogenic and marine isotopic values of total nitrate. The δ15Na was obtained from the mean value of several urban wastewater samples of the region (17.8±0.6‰, n=7). And the δ15Nm is the mean value of deep oceanic water from the Bay of A Coruña (6.02±0.28‰, n=24). Each bar represents the % contribution of both sources estimated using 3 hypothetical isotopic enrichment factors (Ɛ). 151 Chapter 7: General discussion not known, the isotopic values in macroalgae can lead to misinterpretation of the contribution of anthropogenic sources (Fig. 7.2). Because of the higher enrichment values in macroalgae compared to water observed in Chapter 6, the possible influence of fractionation during N uptake can be discarded for these species, as it was shown for other species (Cohen and Fong 2005, García-Sanz 2009). The mechanism causing this positive fractionation in the macroalgae may be related to nutrient release (Umezawa et al. 2007). This process is well studied in animals as is the base of food webs studies (Montoya 2008) but it has been poorly studied in macroalgal species. Although the existence of organic and inorganic nutrient release has been observed in some macroalgae (Tyler et al. 2001, Naldi and Wheeler 2002), little is known about the isotopic fractionation processes or the environmental factors involved. The isotopic values of A. nodosum and Fucus spp. growing at the same site, although correlated, were different. This result suggests the existence of differences in the fate of N taken up for the different species. These differences may be related to the uptake process or, more likely, to the subsequent release of part of the nitrogen after incorporation into the algal tissues. Such differential processing of N among species would have important implications for the use of a particular species as biomonitor of N loadings, especially when comparing results between systems or regions of the same system where species composition differs. However, species differences in nitrogen uptake may be useful for understanding macroalgae responses to anthropogenic sources and elevated N concentrations. From these results we can conclude that the integration times of these species are particularly appropriate to differentiate chronic pollution from point discharges that may have little impact on brown macroalgae. Considering the current limitations highlighted in the studies included in this thesis, future prospects can be disclosed in two main aspects. First, further studies aiming to understand the link between species-specific N metabolism and its link with isotope fractionation are needed to fully interpret the current data. Second, the existence of fractionation during release of internal N needs to be ascertained as it would explain the enrichment of macroalgal tissues relative to the water DIN. 153153 General Conclusions 1. The δ15N values of A. nodosum and F. vesiculosus are significantly influenced by the isotopic signature of both anthropogenic (wastewater) and natural sources (upwelling). The interpretation of results at large geographic scales including a gradient of upwelling requires previous knowledge of the influence of upwelling on δ15N at each site. 2. The δ15N of A. nodosum and F. vesiculosus increase with the size of human population nuclei for populations <15,000 inhabitants but with large associated variability, whereas for larger population sizes the isotopic values are unrelated with the size of the population. This suggests the existence of differences in the efficiency of nitrogen removal from wastewater of small cities and villages or differences among water treatments. 3. The starting populations of F. vesiculosus after experimental denudations were established through the recruitment of three different cohorts during a 17-month period. The growth curves obtained followed a logistic pattern and were different between cohorts established at different times during the year, and between sites (semi-exposed and estuarine). Overall, a 1-cm tip would represent the growth of the previous 14 days for an individual at an estuarine site. 4. The recruitment of A. nodosum in experimentally denudated areas was a slow process and only one cohort was detected during the 26-month study period. When studying adult populations, differences among growth curves estimated for different localities were observed, highlighting the importance of local factors for the growth of this species. Overall, at an estuarine site, a 1-cm tip would approximately represent the individual´s growth of the previous 21 days. 5. The basal and old growth parts of the fronds of A. nodosum and F. vesiculosus cannot be recommended for using in retrospective isotopic studies because they maintain N uptake capabilities during the individual´s life. The solely use of growing tips is recommended instead as they show apical growth and no transport was observed along the thallus. Chapter 8: General conclusions