Portuguese Native Artemia Parthenogenetica and Invasive

*Manuscript Click here to download Manuscript: Pinto et al. manuscript.docx Click here to view linked References

1 1 2 3 2 4 5 3 Artemia parthenogenetica Artemia 6 Portuguese native and invasive 7 8 4 franciscana reproductive parameters, under different abiotic conditions. 9 10 11 5 12 6 Pedro M. Pinto1*; Ana Bio1; Francisco Hontoria3; & Natividade Vieira1,2 13 14 15 7 16 1 CIMAR/CIIMAR – Centre of Marine and Environmental Research, University of 17 18 8 Porto, Portugal, Rua dos Bragas, 289, 4050-123 Porto, Portugal. 19 20 9 2 Department of Biology, Faculty of Sciences, University of Porto, Portugal. Rua do 21 22 23 10 Campo Alegre s/n, 4169-007 Porto, Portugal. 24 25 11 3 Instituto de Acuicultura de Torre de la Sal (IATS - CSIC), 12595 Ribera de Cabanes 26 27 28 12 (Castellón), Spain. 29 30 13 31 32 33 14 *Corresponding author: [email protected] 34 35 36 37 15 38 16 Abstract 39 17 40 Artemia 41 18 There are only two known populations of native in Portugal: one in the 42 43 19 Rio Maior saline, the other in the Aveiro salines complex, both of the diploid Artemia 44 45 20 parthenogenetica species. All other Portuguese hypersaline environments where 46 47 48 21 Artemia can be found have been invaded by Artemia franciscana, which has eradicated 49 50 22 the native strains. Given the actual widespread interest in the conservation of native 51 52 53 23 Artemia biodiversity, the survival of the two Portuguese’ native Artemia strains and of 54 55 24 the invasive A. franciscana, when exposed to variations of several abiotic factors, was 56 57 58 25 recently studied, and the differences in the survival of the Portuguese Artemia species 59 60 26 were evident. To complement that previous study, this work evaluates the reproductive 61 62 63 64 65 27 performance of the two native strains and of the invasive species, when exposed to the 1 2 28 same abiotic variations. After both studies, the A. parthenogenetica from Rio Maior 3 4 5 29 seems to be very well adapted to its specific biotope characteristics, which, together 6 7 30 with the inland saline localisation, favours its resistance to invasion. On the other hand, 8 9 10 31 the A. parthenogenetica from Aveiro proved to perform much worse than its invasive 11 12 32 competitor, in the conditions tested. Its permanence in its biotope is a still unexplained 13 14 33 phenomenon. The only two explanations that we currently glimpse for this fact are that 15 16 17 34 either A. franciscana has by chance not been introduced there, or a chemical barrier 18 19 35 related to the pollution has been preventing invasion. Further studies are needed to 20 21 22 36 discern the true reasons, and they have to take into account relations between local 23 24 37 conditions, such as environmental problems, and the specific biological traits of the 25 26 27 38 local Artemia strains. 28 29 39 30 31 32 40 Keywords: Artemia, biological traits, invasive species, abiotic conditions, salines, 33 34 35 41 Portugal. 36 37 38 42 39 43 40 44 Introduction 41 42 45 43 46 Artemia is a cosmopolitan genus (Pacios and Muñoz, 2010). Nine reproductively 44 45 47 isolated species, with sexual or parthenogenetic reproduction (Browne and Bowen, 46 47 48 48 1991) have been identified, four of which with natural populations on the Iberian 49 50 49 Peninsula (Amat et al., 1995; Amat et al., 2007; Pacios and Muñoz, 2010). There are, 51 52 53 50 however, only two known populations of native Artemia in Portugal: one in the Rio 54 55 51 Maior saline (39 ° 21'47 "N8 ° 56'33" W), the other in the Aveiro saline complex 56 57 58 52 (40°38'37" N, 8°39'57" W), both of the diploid Artemia parthenogenetica species (Amat 59 60 53 et al., 2007; Pinto et al., 2013a). All other Portuguese hypersaline environments where 61 62 63 64 65 54 Artemia can be found have been invaded by Artemia franciscana (Amat et al., 2007, 1 2 55 Pinto et al., 2013a), which has eradicated the native strains. 3 4 5 56 Artemia franciscana is a native species from the American continent (Amat et 6 7 57 al., 2005), but can currently be found on all continents where Artemia has been 8 9 10 58 described, evidencing its large invasive power (Ruebhart et al., 2008). Several studies 11 12 59 (e.g. Ruebhart et al., 2008) describe characteristics of A. franciscana that favour 13 14 60 development and spread of this invasive species in comparison to other Artemia species. 15 16 17 61 A. franciscana is very euryhaline and eurythermal, maintaining reproductive success at 18 19 62 a variety of temperatures and salinities (Browne and Wanigasekera, 2000). Amat et al. 20 21 22 63 (2007), compared the sexual fitness of several Artemia species from different places of 23 24 64 the Mediterranean basin, under standard conditions, and found that A. franciscana had 25 26 27 65 the best reproductive performance. 28 29 66 Given the actual widespread interest in the conservation of native Artemia 30 31 67 32 biodiversity, Pinto et al. (2013b) recently studied the survival of the two Portuguese’ 33 34 68 native Artemia strains and of the invasive A. franciscana, when exposed to variations of 35 36 69 several abiotic factors: salinity, temperature, light and quantity of available food. They 37 38 39 70 found significant differences in the survival of the Portuguese Artemia species, showing 40 41 71 that native Artemia strains vary in terms of resistance to abiotical changes. Distinct 42 43 44 72 physiological responses of different Artemia populations belonging to the same species 45 46 73 are common in this genus (Browne and Bowen, 1991; Browne, 1992). These variations, 47 48 49 74 which are possibly genetic, suggest local adaptations of the populations, in response to 50 51 75 different selective pressures experienced in the most varied hypersaline environments 52 53 76 they inhabit (Persoone and Sorgeloos, 1980; Vanhaecke et al., 1984). Pinto et al. 54 55 56 77 (2013b) suggested possible local adaptations of the Portuguese native Artemia species 57 58 78 to specific characteristics of the studied biotopes (the salines of Aveiro, located in a 59 60 61 62 63 64 65 79 large lagoon, and the inland saline of Rio Maior), and identified local factors which 1 2 80 potentially limit the invasive ability of A. franciscana. However, to fully understand the 3 4 5 81 invasive capacity of A. franciscana in these environments and the resilience of the 6 7 82 native Artemia strains which resist invasion, next to survival, reproductive features and 8 9 10 83 the dynamics of these three populations need to be assessed. This work complements 11 12 84 Pinto et al. (2013b), studying the reproductive performance of the two native strains and 13 14 85 of the invasive species when exposed to abiotic variations in salinity, temperature, light 15 16 17 86 and food availability. Pre-reproductive and reproductive periods, as well as the type and 18 19 87 quality of reproduction are assessed, as these factors decisively influence the biological 20 21 22 88 fitness and lifetime of each of these species (Allan, 1976; Barata et al., 1995, 1996a,b; 23 24 89 Browne et al., 1984, 1988, 1991), as well as the permanence or elimination of native 25 26 27 90 strains in these hypersaline environments (Amat et al., 2007). 28 29 91 30 31 92 Material and methods 32 33 93 34 94 35 95 Biological material 36 96 37 97 Adult diploid A. parthenogenetica were collected at two sites: the Troncalhada 38 39 40 98 saline (40º38’40”N, 8º39’46”W) located in the Ria de Aveiro lagoon, an artisanal, solar 41 42 99 saline covering an area of 4.2 ha; and the saline of Rio Maior (39º21’50”N, 43 44 45 100 8º56’38”W), an inland saline supplied by pumped up brine from naturally dissolved 46 47 101 rock salt (more information about these salines can be found in Pinto et al., 2013b) 48 49 50 102 (Fig. 1). A. franciscana were hatched from a commercial brand of Artemia cysts (Ocean 51 52 103 Nutrition™, Great Salt Lake). The animals were kept in the laboratory, separated per 53 54 55 104 population, and allowed to acclimatise to a salinity of 70 ppt and a temperature of 24C. 56 57 105 The first 24 females of each Artemia strain to reach sexual maturity (or less, 58 59 106 60 when fewer animals reached sexual maturity during the experiment, which happened for 61 62 63 64 65 107 some extreme experimental conditions) were transferred immediately and individually 1 2 108 (for A. parthenogenetica) or together with a male (for A. franciscana) to 50 ml Falcons 3 4 5 109 tubes, maintaining the acclimatization culture conditions. Females were considered to 6 7 110 be sexually mature after developing their ovisac and males when their antenna and penis 8 9 10 111 were clearly observable. 11 12 112 13 14 113 Experimental setup 15 16 17 114 We considered 70 ppt salinity, 24°C water temperature, 12:12 h L:D photoperiod 18 19 115 and 300000 cells ml−1 Tetraselmis suecica food supply as standard conditions. In each 20 21 22 116 experiment one of these parameters was varied to assess its effect on reproduction; the 23 24 117 remaining parameters were kept constant. Treatment conditions were: salinities of 25 26 27 118 70 ppt, 110 ppt and 150 ppt (prepared using natural sea water and Tropic Marin Sea 28 29 119 Salt® and confirmed with a refractometer); temperatures of 24°C, 29°C and 34°C ±1°C 30 31 120 32 (maintained keeping the falcons in water baths, with temperatures regularly checked 33 34 121 with a thermometer); photoperiods of 12:12h L:D, constant light and constant darkness; 35 36 122 and three levels of food supply, with ±300000 cells ml−1, ±150000 cells ml−1 and 37 38 −1 39 123 ±37500 cells ml of T. suecica respectively (densities were obtained through dilution, 40 41 124 counting T. suecica cells in a Neubauer chamber). Notice that supplied food 42 43 44 125 concentrations were independent of the number of individuals present in each replicate 45 46 126 (one female for parthenogenetic populations and a couple for the bisexual species), but 47 48 49 127 the medium was renewed every second day. 50 51 128 Each treatment condition was applied to up to 24 replicates (depending on the 52 53 129 number of mature females obtained). For each replicate, the experiment finished with 54 55 56 130 the death of the parthenogenetic individual or the A. franciscana female. If a male A. 57 58 59 60 61 62 63 64 65 131 franciscana died before its replicate’s female, the male was replaced by another that had 1 2 132 been kept under the same treatment conditions. 3 4 5 133 Every second day, the animals were observed to determine the following life 6 7 134 history features: pre-reproductive, reproductive and post-reproductive periods (in days); 8 9 10 135 life span (days); broods per female; inter-brood interval (days); total offspring; offspring 11 12 136 per brood and female; ovoviviparous offspring percentage and quality; and oviparous 13 14 137 offspring percentage quality. Cyst quality was evaluated by its chorion appearance and 15 16 17 138 floating capacity only; good external appearance and floating capacity in the salinities 18 19 139 tested were considered indicators of good reproductive viability. The time between 20 21 22 140 broods was assessed considering only females that produced at least 3 broods during the 23 24 141 experiment (seen as females with a consistent breeding period). 25 26 27 142 28 143 2.2 Data analysis 29 144 30 145 31 146 32 Means and SD of life periods and offspring data were plotted and compared for 33 34 147 the different Artemia strains and culture conditions. Given that many treatment results 35 36 148 failed the Shapiro-Wilk test for normality (with p<0.05) (Royston, 1982), also when log 37 38 39 149 transformed, differences between Artemia sources and between treatment effects were 40 41 150 assessed through a distribution-free analysis of variance, using the Kruskal-Wallis test 42 43 44 151 to determine if any treatment group was different from the others, and post-hoc pairwise 45 46 152 Wilcoxon rank sum tests, corrected for multiple testing (with p-value adjustment by the 47 48 49 153 Holm method; Holm 1979), to see which groups differed (Hollander and Wolfe, 1973). 50 51 154 Multivariate analyses were used to compare Artemia from different sources. A 52 53 155 correlation-based Principal Component Analysis (PCA; the DCA gradient lengths were 54 55 56 156 all <3) of the life table and reproductive parameters was carried out for each 57 58 157 experimental treatment. The distinction between Artemia strains was further tested 59 60 61 62 63 64 65 158 through an Analysis of Similarity (ANOSIM), following a 2D Non-Metric 1 2 159 Multidimensional Scaling (NMDS) based on standardized and normalized Euclidian 3 4 5 160 distances. Multivariate analyses did not consider time between broods, to avoid 6 7 161 eliminating females that produced less than three broods from the data. 8 9 10 162 PCA were performed using CANOCO 4.5 (TerBraak and Smilauer, 1998), 11 12 163 NMDS and ANOSIM using Primer 5 (Clarke and Warwick, 2001). All other statistical 13 14 164 analyses were done in R (R Dev. Core Team, 2009). 15 16 17 165 18 19 166 Results 20 21 167 22 168 Survival and maturation 23 24 169 25 170 Not all of the experimental conditions lead to the survival (Pinto et al., 2013b) 26 27 28 171 and maturity of 24 females (Table 1). For instance, at 34ºC temperature no animals of 29 30 172 any of the studied Artemia strains reached maturity. And at 29ºC, only 13 individuals 31 32 33 173 from Rio Maior reached maturity and only one produced more than 2 broods. High 34 35 174 salinities proved to be deadly for A. parthenogenetica from Aveiro (only one individual 36 37 38 175 matured at 110 ppt, none at 150 ppt) and A. franciscana (no animal matured at 150 ppt). 39 40 176 A. parthenogenetica from Rio Maior, on the other hand, was more tolerant to high 41 42 177 Artemia 43 salinity. Food shortage affected mainly from Aveiro. The photoperiods studied 44 45 178 hardly affected survival and maturation. 46 47 48 179 49 50 51 180 Life time and reproductive periods 52 53 181 Mean values (and SD) of the Life time and reproductive periods are presented in 54 55 182 Artemia 56 Figure 2. The Wilcoxon rank sum test results comparing strains are presented 57 58 183 in Table 2, and those comparing treatment levels for each strain in Table 3. 59 60 61 62 63 64 65 184 Under standard conditions, A. parthenogenetica from Rio Maior had the longest 1 2 185 mean pre-reproductive period (PRP) of all tested species; the shortest was obtained for 3 4 5 186 A. franciscana. Both A. franciscana and A. parthenogenetica from Rio Maior showed 6 7 187 the longer mean reproductive periods (RP) than Artemia from Aveiro, which presented 8 9 10 188 a relative long PRP, but short RP and total life span (LS) (Fig. 2). 11 12 189 At 110 ppt salinity, the mean PRP of A. franciscana was significantly shorter 13 14 190 than that of Artemia from Rio Maior, and the RP was bigger than at 70 ppt for all 15 16 17 191 species, although this difference was only significant for the A. franciscana. Too few 18 19 192 (1) animals of A. parthenogenetica from Aveiro survived to allow comparison. At 150 20 21 22 193 ppt only Artemia from Rio Maior was able to reproduce. 23 24 194 At 29ºC temperature, A. parthenogenetica from Aveiro showed significantly 25 26 27 195 longer PRP and (consequently) LS than the other strains and that at 24ºC. Both A. 28 29 196 parthenogenetica from Rio Maior and A. franciscana reduced their RP and LS at 29ºC, 30 31 197 Artemia 32 in comparison to the 24ºC treatment. from Rio Maior had the shortest RP. At 33 34 198 29ºC A. franciscana displayed shorter pre and post reproductive periods than the strain 35 36 199 from Aveiro. 37 38 39 200 Halving the food supply strongly affected A. parthenogenetica from Aveiro, 40 41 201 lengthening its PRP and shortening (though not significantly) its RP. This strain did not 42 43 −1 44 202 survive extreme food shortage (37500 cells ml of T. suecica). Differences between the 45 46 203 less affected Artemia form Rio Maior and A. franciscana were non-significant for the 47 48 49 204 halved food supply, but significant under extreme food shortage, when the RP of 50 51 205 Artemia from Rio Maior became very short. 52 53 206 Continuous light caused shorter PRP and LS for A. franciscana, in comparison 54 55 56 207 to the other strains. Continuous darkness mainly decreased the RP and LS of A. 57 58 208 parthenogenetica from Rio Maior. 59 60 61 62 63 64 65 209 1 2 210 Broods and offspring quantity 3 4 5 211 Mean values (and SD) of brood and offspring numbers are presented in Figure 3. 6 7 212 The Wilcoxon rank sum test results comparing Artemia strains are presented in Table 2, 8 9 10 213 and those comparing treatment levels for each strain in Table 3. 11 12 214 Under standard conditions, there were significantly less broods per female 13 14 215 (BPF), offspring per female (OPF) and offspring per brood (OPB) for A. 15 16 17 216 parthenogenetica from Aveiro than for the other two strains. A. franciscana had 18 19 217 furthermore more OPB and a shorter time between broods (TBB) than Artemia from 20 21 22 218 Rio Maior. At 110 ppt, A. franciscana produced significantly more OPF and OPB than 23 24 219 Artemia from Rio Maior, and more BPF and OPF than under standard conditions. At 25 26 27 220 150 ppt only A. parthenogenetica from Rio Maior was able to reproduce. An increase in 28 29 221 temperature to 29ºC strongly reduced the number of BPF and OPF in A. franciscana 30 31 222 Artemia A. parthenogenetica 32 and from Rio Maior. from Rio Maior had the lowest 33 34 223 reproduction, though most differences between strains were not significant. Halving the 35 36 224 food supply reduced OPF and OPB in A. franciscana and Artemia from Aveiro, but did 37 38 39 225 not negatively affect Artemia from Rio Maior (except for the OPB). Artemia from 40 41 226 Aveiro produced significantly less BPF, OPF and OPB than the other two strains. Under 42 43 44 227 extreme food shortage, only A. franciscana produced few but still more BPF and OPF 45 46 228 than Artemia from Rio Maior. Under both continuous light and continuous darkness, 47 48 Artemia A. franciscana 49 229 from Rio Maior produced less BPF and OPF. preferred 50 51 230 continuous light to continuous darkness, whereas Artemia from Aveiro showed little 52 53 231 photoperiod effects. 54 55 56 232 There were few significant differences in time between broods (TBB), 57 58 233 comparing Artemia strains or treatment levels. Under standard conditions and at 110 ppt 59 60 61 62 63 64 65 234 salinity, A. franciscana had shorter TBB than Artemia from Rio Maior, but both strains 1 2 235 did not differ in TBB from the Aveiro strain. Under continuous light A. franciscana had 3 4 5 236 shorter TBB than the other strains. 6 7 237 8 9 10 238 Offspring type and quality 11 12 239 Offspring type and quality are quantified in Figure 4, as mean percentages per 13 14 240 Artemia strain and experimental condition. Mean values (and SD), as well as Wilcoxon 15 16 17 241 rank sum test results comparing Artemia strains in terms of offspring quantity and 18 19 242 quality are presented in Table 4. 20 21 22 243 Under standard conditions, the offspring from A. parthenogenetica from Rio 23 24 244 Maior and Aveiro was about half ovoviviparous, but offspring quantity and quality was 25 26 27 245 much worse for the Aveiro strain (Fig. 4, Table 4). A. franciscana reproduced mainly 28 29 246 through good quality oviparous offspring. An increase in salinity (to 110 ppt), 30 31 247 32 temperature (29ºC) or light (24h) led to predominantly ovoviviparous reproduction and 33 34 248 less offspring quality for the Aveiro strain, while food shortage caused more oviparous 35 36 249 offspring. Artemia from Rio Maior, on the other hand, reacted with predominantly 37 38 39 250 oviparous reproduction to 110 ppt salinity and continuous light, but with ovoviviparous 40 41 251 reproduction under food shortage and at 29ºC (with many dead and abortive embryos). 42 43 44 252 At 150 ppt only Artemia from Rio Maior was able to reproduce, mostly with good 45 46 253 quality offspring. A. franciscana produced more cysts than the other strains at 29ºC and 47 48 49 254 under food shortage, but with rather bad quality offspring when food shortage was 50 51 255 extreme. Compared to standard conditions, continuous light or darkness and increased 52 53 256 temperature (except for A. franciscana) caused losses in offspring quality. 54 55 56 257 57 58 258 Multivariate analysis 59 60 61 62 63 64 65 259 Considering base conditions the first axis of the PCA separates samples 1 2 260 according to offspring quantity and the second axis according to offspring quality 3 4 5 261 (Fig. 5). Artemia from Aveiro are characterised by less offspring and broods, whereas 6 7 262 Artemia from Rio Maior and A. franciscana samples are separated on the second axis, 8 9 10 263 according to their preference for oviparous and ovoviviparous reproduction, 11 12 264 respectively. 13 14 265 At 110 ppt salinity only one Artemia from Aveiro matured. Again Artemia from 15 16 17 266 Rio Maior and A. franciscana samples are separated on the second axis, according to 18 19 267 their main reproduction mode, but also due to the longer pre-reproductive periods and 20 21 22 268 less OPB observed for Artemia from Rio Maior. 23 24 269 High temperature lead to a well defined cluster of Artemia from Rio Maior, on 25 26 27 270 the left side of the first axis, with less offspring, broods and ovoviviparous reproduction. 28 29 271 Artemia from Aveiro and A. franciscana are separated on the second PCA axis, 30 31 272 32 characterised by less oviparous reproduction and longer PRP and post-reproductive 33 34 273 periods in samples from Aveiro. 35 36 274 Halving the food supply showed similar patterns as the base condition: a first 37 38 39 275 axis separating samples according to offspring quantity and a second one according to 40 41 276 offspring quality, with Artemia from Aveiro characterised by less offspring and broods, 42 43 44 277 and Artemia from Rio Maior and A. franciscana separated, according to their preference 45 46 278 for oviparous and ovoviviparous reproduction, respectively. Samples from Aveiro show 47 48 Artemia A. 49 279 furthermore long PRP. Extreme food shortage separates from Rio Maior and 50 51 280 franciscana on the first axis, with less offspring and broods, less oviparous offspring 52 53 281 and longer PRP for samples from Rio Maior. 54 55 56 282 Continuous light condition separated Artemia from Rio Maior on the first axis, 57 58 283 with samples characterised by less offspring and broods, and less ovoviviparous 59 60 61 62 63 64 65 284 offspring, and on the second axis with samples characterised by less oviparity. 1 2 285 Complete darkness also separated Artemia from Rio Maior on the first axis due to less 3 4 5 286 offspring and broods, and due to longer PRP. Under these conditions, Artemia from 6 7 287 Aveiro and A. franciscana were mainly separated on the second axis, characterized by 8 9 10 288 the preference of oviparous reproduction in the Aveiro samples. 11 12 289 Analysis of similarity (ANOSIM) after 2D NMDS showed significant (p<0.05) 13 14 290 discrimination between Artemia sources, except for the very overlapping samples of 15 16 17 291 Artemia from Rio Maior and A. franciscana kept at 110 ppt salinity, and often little 18 19 292 separation between samples from different sources (R<0.250). Reasonable separation 20 21 22 293 was obtained for the Rio Maior samples in continuous light (R>0.500) and food 23 24 294 shortage conditions (R>0.400), and between the Rio Maior and Aveiro samples kept in 25 26 27 295 complete darkness (R>0.400). 28 29 296 30 31 297 Discussion 32 33 34 298 The results of the present study, together with those obtained previously by Pinto 35 36 299 et al. (2013b), provide a broad understanding of the different responses of the two 37 38 39 300 studied Portuguese native parthenogenetic Artemia strains and the invasive A. 40 41 301 franciscana to environmental conditions. They also provide clues to species 42 43 44 302 invasiveness or resistance to invasion and guidelines for the best abiotic conditions for 45 46 303 each of these Artemia species. 47 48 49 304 The native parthenogenetic strains from Aveiro and Rio Maior showed very 50 51 305 different reproductive performance, in quantity and in quality, confirming the variability 52 53 306 in diploid Artemia strains (Pinto et al., 2013b; Browne, 1992). A. franciscana showed 54 55 56 307 overall better survival (Pinto et al., 2013b) and reproductive abilities and, therefore 57 58 308 potential competitive advantage over A. parthenogenetica from Aveiro. This advantage 59 60 61 62 63 64 65 309 was less evident compared to the Rio Maior strain; this strain even survived and 1 2 310 reproduced better than A. franciscana at high salinities (150 ppt) and with halved food 3 4 5 311 supply. However, available food density was the same for the single parthenogenetic 6 7 312 Artemia and the couple of A. franciscana in each experiment (although the medium was 8 9 10 313 renewed every second day), which favoured the Rio Maior strain. 11 12 314 Previous studies have confirmed the eurythermal characteristic of A. franciscana 13 14 315 (Wear and Haslett, 1986; Browne, 1988; Browne et al., 1991; Browne and 15 16 17 316 Wanigasekera, 2000). A. franciscana showed much higher survival at 29ºC than the 18 19 317 native Portuguese parthenogenetic strains, and even some survival at 34ºC (Pinto et al., 20 21 22 318 2013b). In terms of reproductive ability however, the present study shows that A. 23 24 319 franciscana prefers a lower temperature. This species showed larger reproductive 25 26 27 320 periods and oviparity output and quality at 24ºC than at 29ºC, and animals died without 28 29 321 reproducing at 34ºC. In comparison to the parthenogenetic strains, A. franciscana 30 31 322 32 showed a small advantage at 29ºC, presenting smaller pre-reproductive periods and 33 34 323 more abundant good cysts. Even under base conditions (24ºC, 70 ppt, abundant food 35 36 324 and 12:12 L:D photoperiod), A. franciscana presented overall better reproductive 37 38 39 325 success than A. parthenogenetica from Aveiro and similar success to A. 40 41 326 parthenogenetica from Rio Maior. 42 43 44 327 A. franciscana showed also better adaption to intermediate/high salinities than 45 46 328 the Portuguese parthenogenetic strains, in terms of both survival (Pinto et al., 2013b), 47 48 49 329 and offspring quantity, although it had no clear advantage in terms of offspring 50 51 330 composition and quality. A. franciscana was the only strain showing better reproductive 52 53 331 figures at 110 ppt than at 70 ppt, whereas A. parthenogenetica from Rio Maior 54 55 56 332 performed better at 150 ppt than the other two strains, producing viable offspring at this 57 58 333 high salinity. 59 60 61 62 63 64 65 334 A. parthenogenetica from Rio Maior also performed best in the halved food 1 2 335 supply experiment. Although A. franciscana and A. parthenogenetica from Rio Maior 3 4 5 336 showed both a good adaptation to lack of food, there was a small advantage of the 6 7 337 parthenogenetic strain, which had better quality ovoviviparous and oviparous offspring 8 9 10 338 than A. franciscana (notice that, as mentioned, food density was not the same per 11 12 339 individual, which may have biased this result). This resistance to food shortage had 13 14 340 already been demonstrated in terms of survival (Pinto et al., 2013b). Under extreme 15 16 17 341 food shortage (an eighth of the base food density) A. franciscana and A. 18 19 342 parthenogenetica from Rio Maior still reproduced, but with great difficulty. A. 20 21 22 343 parthenogenetica from Aveiro was the most sensitive to food shortage reproducing 23 24 344 badly at halved food and not even surviving extreme food shortage. 25 26 27 345 In relation to the photoperiods studied, continuous light and continuous darkness 28 29 346 affected reproductive success of Artemia from Rio Maior negatively, A. franciscana 30 31 347 Artemia 32 preferred continuous light to continuous darkness, whereas from Aveiro 33 34 348 showed little photoperiod effects. Although these results, which are mostly in agreement 35 36 349 with those obtained in previous survival experiments, are difficult to relate to natural 37 38 39 350 conditions, where continuous light or darkness never occur, variations in the 40 41 351 reproductive behaviour of these species in changing photoperiods might be important 42 43 44 352 for the invasive capacity of A. franciscana. 45 46 353 So, what has (so far) prevented the introduction A. franciscana in the two 47 48 49 354 studied Portuguese hypersaline biotopes? The Rio Maior saline has unique 50 51 355 characteristics that reduce the possibility of an accidental introduction of the invader. 52 53 356 Firstly, its geographical inland location, which lays far from the main bird migration 54 55 56 357 routes and is far from fish farming facilities and urban areas with aquaria (Amat et al., 57 58 358 2007). Other characteristics that may help reduce the invader introduction are its source 59 60 61 62 63 64 65 359 of saltwater, pumped up from a rock salt mine, with an output salinity of 150 ppt and a 1 2 360 low abundance and diversity of Artemia food organisms (the tanks are made of cement 3 4 5 361 and cleaned at the end of every salt season production, showing little algae 6 7 362 concentration during the salt production season). Finally, the rather deep water tanks 8 9 10 363 used for water storage, which are unusual for salines. The results here presented suggest 11 12 364 that the local A. parthenogenetica is better adapted to the particular conditions of this 13 14 365 saline than A. franciscana. A. parthenogenetica from Rio Maior was the only strain that 15 16 17 366 survived and reproduced at 150 ppt. It had better reproductive results at 24ºC and worse 18 19 367 at 29ºC than its invasive competitor, which may be advantageous in the deep, less 20 21 22 368 warmed up tanks. Artemia from Rio Maior also outperformed A. franciscana when food 23 24 369 supply was halved. But we also found features that should enhance invasiveness of A. 25 26 27 370 franciscana in the Rio Maior biotope: A. franciscana outcompetes the local 28 29 371 parthenogenetic Artemia at very low food concentrations, at higher temperatures and in 30 31 372 32 complete darkness, which may constitute an advantage in the deep tanks of this saline 33 34 373 with little light incidence. 35 36 374 For the native population from the Aveiro salines complex the situation is 37 38 39 375 completely different, both in terms of biotope characteristics and of the weak 40 41 376 reproductive capacity of this strain in comparison to A. franciscana. The Aveiro saline 42 43 44 377 complex is fed by water from the Ria de Aveiro lagoon, with up to 35 ppt salinity. 45 46 378 Water tanks are shallow with natural sediment bottoms. Salinity increases gradually 47 48 49 379 from the inlet channel to the crystallisers, where 300 ppt can be reached (Vieira, 1989). 50 51 380 Local conditions may explain why A. franciscana outcompeted the local A. 52 53 381 parthenogenetica in all experimental conditions. For instance, the weak performance of 54 55 56 382 Artemia from Aveiro at low food concentrations may be due to its adaptation to the 57 58 383 abundant and diversified food species found in the Aveiro saline water (Vieira and Bio, 59 60 61 62 63 64 65 384 2011). Hence, this saline must be seriously threatened by invasion, especially given that 1 2 385 there are records of A. franciscana in a saline of the same complex, not too distant from 3 4 5 386 the saline here studied, and given that the lagoon and salines are inhabited by numerous 6 7 387 bird species that could disseminate the invasive Artemia strain (Amat et al., 2007). The 8 9 10 388 fact that, in spite of its poor survival, biological fitness and lifespan, A. 11 12 389 parthenogenetica persists in the saline may be due to two explanations. First, although 13 14 390 very unlikely for the reasons previously exposed, A. franciscana has not yet reached 15 16 17 391 that saline, which would mean that its introduction should urgently be prevented to 18 19 392 avoid elimination of the native Artemia strain. Second, A. franciscana already appeared 20 21 22 393 in that biotope but was not able to establish itself there. This second option could be 23 24 394 explained by the existence of a chemical barrier, as suggested by Pinto et al. (2013b), 25 26 27 395 related to contaminants in the polluted Ria de Aveiro lagoon water, such as heavy 28 29 396 metals or pesticides (e.g. Martins et al., 2010). Future studies should assess the effect of 30 31 397 Artemia 32 the chemicals present in this particular saline on survival and reproduction, as 33 34 398 local chemical biotope characteristics may be of crucial importance for the maintenance 35 36 399 of native Artemia strains. 37 38 39 400 40 41 401 Conclusions 42 43 44 402 A broad assessment of strain-specific traits is needed to understand competition, 45 46 403 resilience and resistance of native Artemia strains in relation to the invasive A. 47 48 franciscana A. parthenogenetica 49 404 . from Rio Maior seems to be very well adapted to its 50 51 405 specific biotope characteristics, which, together with its inland localisation, favours its 52 53 406 resistance to invasion. It is nevertheless important to take measures to avoid the 54 55 56 407 introduction of A. franciscana, which is probably capable of displacing the local 57 58 408 Artemia strain in biological terms. 59 60 61 62 63 64 65 409 A. parthenogenetica from Aveiro, on the other hand, proved to perform much 1 2 410 worse than its invasive competitor, in the conditions tested. Its permanence in its 3 4 5 411 biotope is still an unexplained phenomenon. The only two explanations that we 6 7 412 currently glimpse for this fact are that either A. franciscana has by chance not been 8 9 10 413 introduced there, or a chemical barrier related to the pollution has been preventing 11 12 414 invasion. Further studies are essential to discern the true reasons. 13 14 415 Future studies are needed to continue to analyse the specific characteristics of 15 16 17 416 the European hypersaline biotopes, taking into account local conditions such as 18 19 417 environmental problems, and relate them to the specific biological traits of the local 20 21 22 418 Artemia strains we aim to preserve, especially against exotic invasive species. 23 24 419 25 26 Acknowledgments 27 420 28 29 421 This study was supported by the FCT (Portuguese Foundation for Science and 30 31 422 Technology) and European funds (FEDER), through the project "Chemical Wars: the role of 32 33 423 chemically mediated interactions in the invasiveness potential of non-native Artemia", 34 35 424 PTDC/MAR/108369/2008 (FCT). This research was partially supported by the European 36 37 38 425 Regional Development Fund (ERDF) through the COMPETE - Operational Competitiveness 39 40 426 Programme and national funds through FCT – Foundation for Science and Technology, under 41 42 427 the project PEst-C/MAR/LA0015/2013. 43 44 45 46 47 48 49 428 50 51 429 52 53 430 54 55 56 431 References 57 58 432 Allan, D., 1976. Life history patterns in zooplankton. Am. Nat. 110, 165–180. 59 60 433 61 62 63 64 65 434 Amat, F., Barata, C., Hontoria, F., Navarro, J.C., Varó, I., 1995. Biogeography of the genus Artemia 1 2 435 (Crustacea, Branchiopoda, Anostraca) in Spain. Int. J Salt Lake Res. 3, 175–190. 3 4 436 5 6 437 Amat, F., Hontoria, F., Ruiz, O., Green, A., Sánchez, M., Figuerola, J., Hortas, F., 2005.The American brine 7 8 438 shrimp as an exotic invasive species in the western Mediterranean. Biol. Invasion 7 (1), 37–47. 9 10 439 11 12 13 440 Amat, F., Hontoria, F., Navarro, J.C., Vieira, N., Mura, G., 2007. Biodiversity loss in the genus Artemia in 14 15 441 the Western Mediterranean Region. Limnetica. 26 (2), 387–404. 16 17 442 18 19 443 Barata, C., Hontoria, F., Amat, F., 1995. Life history, resting egg formation, and hatching may explain the 20 21 444 temporal-geographical distribution of Artemia strains in the Mediterranean basin. 22 23 445 Hydrobiologia 298, 295–305. 24 25 26 446 27 28 447 Barata, C., Hontoria, F., Amat, F., Browne, R., 1996a. Competition between sexual and parthenogenetic 29 30 448 Artemia: temperature and strain effects. J. Exp. Mar. Biol. Ecol. 196, 313–328. 31 32 449 33 34 450 Barata, C., Hontoria, F., Amat, F., Browne, R., 1996b. Demographic parameters of sexual and 35 36 451 parthenogenetic Artemia: temperature and strain effects. J. Exp. Mar. Biol. Ecol. 196, 329–340. 37 38 452 39 40 41 453 Browne, R., 1988. Ecological and genetic divergence of sexual and asexual brine shrimp (Artemia) from 42 43 454 the Mediterranean basin. Nat. Geo. Res. 4, 548–554. 44 45 455 46 47 456 Browne, R., 1992. Population genetics and ecology of Artemia: insights into parthenogenetic 48 49 457 reproduction. Trends Ecol. Evol. 7, 232–237. 50 51 458 52 53 54 459 Browne, R., Bowen, S., 1991.Taxonomy and population genetics of Artemia. Artemia Biology. Browne, R. 55 56 460 A., Sorgeloos, P. &Trotman, C. N. A. Boca Raton, Florida. CRC 9, 221–235. 57 58 461 59 60 462 Browne, R., Wanigasekera, G., 2000. Combined effects of salinity and temperature on 61 62 63 64 65 463 survival and reproduction of five species of Artemia. J. Exp. Mar. Biol. Ecol. 244 (1), 29–44. 1 2 464 3 4 465 Browne, R., Salle, S., Grosch, D., Segreti, W., Purser, S., 1984. Partitioning genetic and environmental 5 6 466 components of reproduction and lifespan in Artemia. Ecology 65, 949–960. 7 8 467 9 10 468 Browne, R., Davis, L., Sallee, S., 1988. Effects of temperature and relative fitness of sexual and asexual 11 12 13 469 brine shrimp Artemia. J. Exp. Mar. Biol. Ecol. 124, 1–20. 14 15 470 16 17 471 Browne, R., Li, M., Wanigasekera, G., Simoneck, S., Brownlee, D., Eiband, G., Cowan, J., 1991. Ecological, 18 19 472 physiological and genetic divergence of sexual and asexual (diploid and polyploid) brine shrimp 20 21 473 (Artemia). Advances Ecol. 1, 41–52. 22 23 474 24 25 26 475 Clarke, K. R., Warwick, R. M., 2001. Change in Marine Communities: an approach to statistical analysis

27 nd 28 476 and interpretation. 2 edition, PRIMER-E Ltd, Plymouth Marine Laboratory, Plymouth. 29 30 477 31 32 478 Hollander, M. and Wolfe, D. A., 1973. Nonparametric Statistical Methods. New York: John Wiley & Sons. 33 34 479 35 36 480 Holm, S., 1979. A simple sequentially rejective multiple test procedure. Scandinavian Journal of 37 38 481 Statistics, 6, 65–70. 39 40 41 482 42 43 483 Martins, V., Ferreira da Silva, E., Sequeira, C., Rocha, F., Duarte, A.C., 2010.Evaluation of the ecological 44 45 484 effects of heavy metals on the assemblages of benthic foraminifera of the canals of Aveiro 46 47 485 (Portugal). Estuarine Coastal Shelf Sci. 87 (2), 293–304. 48 49 486 50 51 487 Pacios, F. and Muñoz, J., 2010. Global Biodiversity and Geographical Distribution Of Diapausing Aquatic 52 53 54 488 Invertebrates: The Case Of The Cosmopolitan Brine Shrimp, Artemia (Branchiopoda, Anostraca). 55 56 489 Crustaceana, 83(4): 465-480. 57 58 490 59 60 61 62 63 64 65 491 Persoone, G., Sorgeloos, P., 1980. General aspects of biogeography of Artemia. In: Persoone, G., Roels, 1 2 492 O., Jaspers, E. (Eds.), The Brine Shrimp Artemia, Ecology, Culturing, Use in Aquaculture, 3. 3 4 493 Universa Press, Wetteren, Belgium, pp. 3–24. 5 6 494 7 8 495 Pinto, P., Amat, F., Almeida, V., Vieira, N., 2013a. Review of the biogeography of Artemia Leach, 1819 9 10 496 (Crustacea: Anostraca) in Portugal. International Journal of Artemia Biology, 3(1): 51-56. 11 12 13 497 14 15 498 Pinto, P., Bio, A., Hontoria, F., Almeida, V., Vieira, N., 2013b. Portuguese native Artemia 16 17 499 parthenogenetica and Artemia franciscana survival under different abiotic conditions. Journal 18 19 500 of Experimental Marine Biology and Ecology. 440, 81–89. 20 21 501 22 23 502 R Development Core Team, 2009. R: A language and environment for statistical computing. R 24 25 26 503 Foundation for Statistical Computing, Vienna, Austria, ISBN 3-900051-07-0, http://www.R- 27 28 504 project.org. 29 30 31 505 Royston, P., 1982. An extension of Shapiro and Wilk's W test for normality to large samples. Applied 32 33 506 Statistics. 31,115–124. 34 35 36 37 507 Ruebhart, D., Cock, I., Shaw, G., 2008. Invasive character of the brine shrimp Artemia franciscana Kellogg 38 39 508 1906 (Branchiopoda: Anostraca) and its potential impact on Australian inland hypersaline 40 41 509 waters. Mar. Freshw. Res. 59 (7), 587–595. 42 43 510 44 45 511 TerBraak, C.J.F.; Smilauer, P., 1998. CANOCO Reference Manual and User's Guide to Canoco for 46 47 512 Windows: Software for Canonical Community Ordination (version 4). Microcomputer Power, 48 49 50 513 Ithaca, NY USA. 51 52 514 53 54 515 Vanhaecke, P., Siddall, S., Sorgeloos, P., 1984. International study on Artemia. XXXII. Combined effects of 55 56 516 temperature and salinity on the survival of Artemia of various geographical origin. J. Exp. Mar. 57 58 517 Biol. Ecol. 80, 259–275. 59 60 518 61 62 63 64 65 519 Vieira, M., 1989. Contribuição para o conhecimento da biologia da Artemia sp. proveniente das salinas 1 2 520 de Aveiro. Sua importância em aquacultura e na dinâmica daquele ecossistema. In Ph.D thesis. 3 4 521 Faculdade de Ciências, Universidade do Porto, Departamento de Zoologia e Antropologia. 5 6 522 7 8 523 Vieira, N., Bio, A., 2011. Spatial and temporal variability of water quality and zooplankton in an artisanal 9 10 524 salina. J. Sea Res. 65, 293–303. 11 12 13 525 14 15 526 Wear, R., Haslett, S., 1986. Effects of temperature and salinity on the biology of Artemia franciscana 16 17 527 (Kellogg) from Lake Grassmere, New Zealand.1 growth and mortality.J. Exp. Mar. Biol. Ecol. 98, 18 19 528 153–166. 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Table(s) Click here to download Table(s): Pinto et al. tables.docx

1 Table 1. Total number of mature females (obtained from the initial 30) and number of females producing 2 more than two broods obtained for the different treatment conditions. The first treatment constitutes the 3 base treatment. Treatments used to make comparisons per treatment variable are marked in bold. A. parthenogenetica females Treat- Temper- Food Salinity Light Rio Maior Aveiro A. franciscana ment aturea supply total >2 broods total >2 broods total >2 broods 1 70 24 12 300000 24 20 22 11 23 21 2 110 24 12 300000 12 9 1b 1b 24 23 3 150 24 12 300000 18 9 0b 0b 0b 0 4 70 29 12 300000 13 1b 18 11 22 9 5 70 24 12 150000 23 16 23 4 24 20 6 70 24 12 37500 12 0b 0b 0b 14 3 7 70 24 0 300000 24 3 23 18 21 19 8 70 24 24 300000 24 4 23 11 24 12 4 a No animals reached maturity at 34ºC 5 b Too few samples to allow statistical analysis 6 7 8 Table 2. Pairwise Wilcoxon rank sum test results for the comparison of lifetable parameters between 9 Artemia from different sources (AV: A.parthenogenetica from Aveiro, RM: A.parthenogenetica from 10 Rio Maior, FR: A. franciscana, c./ml: cells ml1 of Tetraselmis food supply), considering the different 11 treatments, using the Holm (1979) adjustment for multiple comparisons; —: not enough data for testing, 12 ns: non-significant, *: =0.05, **: =0.01, ***: =0.001. Only females producing at least three broods 13 were considered for the calculation of the time between broods.

 1

 1 broods 

female brood

Treatment Source

embryos

female

cysts embryos

reproductive embryos cysts - reproductive

- Pre Reproductive Post Lifespan Broods between Time Offspring Offspring ovoviviparous Life Dead Abortive oviparous Good Bad

Base AV-RM ** *** ns *** *** ns *** *** ** ** ns ns ** ** ns AV-FR *** *** ns * *** ns *** *** ns ns ns ns *** *** * RM-FR *** ns ** * ns ** ns ** * * ns ns ** * * 110 ppt AV-RM — — — — — — — — — — — — — — — AV-FR — — — — — — — — — — — — — — — RM-FR *** ns ns ns ns * * ** ns ns ns ** ns ns ns 150 ppt AV-RM — — — — — — — — — — — — — — — AV-FR — — — — — — — — — — — — — — — RM-FR — — — — — — — — — — — — — — — 29ºC AV-RM *** * ns *** * ns * ns ns * ns ns ns ns ns AV-FR *** ns * *** ns ns ns * ns ns *** ns ns *** ns RM-FR *** * ns ns ns ns * * ns ns ns ns * *** ns 150000 c./ml AV-RM *** *** *** ns *** ns *** *** *** *** ns ns * ** ns AV-FR *** *** ns ns *** ns *** *** ns ns ns ns *** *** ns RM-FR ns ns ns ns ns ns ns ns *** *** ns ** *** * ns 37500 c./ml AV-RM — — — — — — — — — — — — — — — AV-FR — — — — — — — — — — — — — — — RM-FR *** *** * ns *** — *** ns ns *** — * *** ** *** 24h AV-RM ns ns * ** * ns ns ** *** *** ** *** *** *** *** AV-FR *** ns ns ** ns * ns ** * ns ns ns * ** ns RM-FR *** ns ns ** * * ** ns *** *** ** ** *** *** *** 0h AV-RM *** *** ns * *** ns ** ns *** *** ns ns *** ** ** AV-FR ** ns ns ns * ns *** *** *** *** ** *** ns ns ns RM-FR *** *** ns * *** ns *** *** *** *** * *** * ns ns 14

15 16 Table 3. Pairwise Wilcoxon rank sum test results for the comparison of lifetable parameters between 17 treatment levels, considering Artemia from different sources (AV: A.parthenogenetica from Aveiro, RM: 18 A.parthenogenetica from Rio Maior, FR: A. franciscana, c./ml: cells ml1 of Tetraselmis food supply), 19 using the Holm (1979) adjustment for multiple comparisons; —: not enough data for testing, ns: 20 non-significant, *: =0.05, **: =0.01, ***: =0.001. Only females producing at least three broods were 21 considered for the calculation of the time between broods.

 1



Source Parameter

reproductive - reproductive - Pre Reproductive Post Lifespan Broodsfemale betweenbroodsTime Offspringfemale Offspringbrood ovoviviparous Lifeembryos Deadembryos Abortiveembryos oviparous Goodcysts Badcysts

AV 70-110ppt ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 24-29C *** ns * *** ns ns ns * ns ns ** ns ns ** ns

300000-150000c./ml *** ns * ** ns ns * ** *** *** ns ns ns ns ns

12-24 h ** ns ns * ns ns ns ns * * ns * ns ns ns

12-0 h ** * ns *** ns ns ns ns *** *** ns ns *** ** **

24-0 h ns ns ns * ns ns ns ns *** *** ns *** *** *** ***

RM 70-110ppt *** ns ns ns ns ** ns ns * * ns ns ns ns * 70-150ppt ** * ns ns ** ns *** *** * * ns * *** *** ns

110-150ppt ns ns ns ns ns * * *** ns ns ns ns * * **

24-29C ns *** ns *** *** ns *** *** *** *** ns ns *** *** ns

300000-150000 c./ml *** ns ns ns ns — ns * ns ns ns ns ns ns ns

300000-37500 c./ml *** *** ns *** *** — *** *** *** *** ** ns *** *** **

150000-37500 c./ml *** *** ns ns *** — *** *** *** ** * * *** *** **

12-24 h *** *** *** *** *** ns *** ns *** *** ** ns ns ns ***

12-0 h *** *** ns * *** ns *** ns ** *** ns ns * * ns

24-0 h *** ns ns *** ns ns ns ns ** ** ns ns * ns ***

FR 70-110ppt *** *** ** *** *** *** *** ns *** ** ns *** ns ns ns 24-29C ns *** ns ** *** ns *** *** ns ns ns ns *** *** *

300000-150000 c./ml *** * ns ns * ns *** *** ns ns ns ns * * ns

300000-37500 c./ml *** ** *** *** *** ns *** *** ns ns ns * *** *** ns

150000-37500 c./ml *** ** *** *** ** ns *** *** ns ns ns ** *** *** ns

12-24 h ** *** ns ** *** ns *** ns ** ns ns ** *** *** ***

12-0 h *** ns ns * ns ns ns ns ** ns ns *** * ** ns

24-0 h ns *** ns *** *** ns *** ns ns ns ns * *** *** * 22 23 24 Table 4. Mean offspring quantities (and SD) for the different Artemia strains (AV: A. parthenogenetica 25 from Aveiro, RM: A. parthenogenetica from Rio Maior, AF: A. franciscana), depending on treatment 26 conditions; emb.: embryos. Different superscript letters indicate significant differences between Artemia 27 source according to the multiple Wilcoxon rank sum tests;* means that only one female did reproduce, so 28 that no statistical testing was possible. 29 Treat. Source Ovoviviparous Life embryos Dead emb. Abortive emb. Oviparous Good cysts Bad cysts a a a a a a a Base AV 80.6 (80.7) 67.5 (66.9) 6.3 (21.5) 6.8 (20.9) 92.2 (167.2) 49.3 (69.8) 42.9 (120.3) RM 252.0 (232.1)b 245.6 (228.4)b 4.4 (9.4)a 2.0 (5.5)a 225.8 (360.1)b 201.1 (351.1)b 24.8 (51.6)a FR 162.8 (279.2)a 158.3 (275.7)a 3.6 (10.5)a 0.9 (3.0)a 484.3 (402.8)c 387.5 (375.0)c 96.7 (103.4)b

110 AV 321.0 (0.0)* 311.0 (0.0)* 4.0 (0.0)* 6.0 (0.0)* 11.0 (0.0)* 0.0 (0.0)* 11.0 (0.0)* Ppt RM 96.0 (40.6)a 90.1 (40.9)a 1.5 (1.4)a 4.4 (6.3)a 507.7 (513.1)a 302.4 (302.1)a 205.3 (240.5)a a a a b a a a FR 605.7 (617.3) 397.5 (397.9) 1.9 (2.8) 206.4 (300.5) 671.4 (556.8) 492.0 (472.4) 179.3 (240.4)

150 AV

Ppt RM 92.7 (51.1) 83.6 (49.4) 1.2 (2.4) 7.9 (9.5) 38.4 (68.6) 24.8 (39.5) 13.6 (31.3) FR

a a a a ab a a 29ºC AV 109.4 (105.1) 98.7 (96.6) 2.9 (3.8) 7.8 (14.2) 22.2 (21.1) 0.2 (0.7) 22.1 (20.9) a b ab a b a a RM 35.9 (21.0) 23.8 (17.6) 2.5 (3.4) 9.7 (19.8) 6.8 (9.9) 0.0 (0.0) 6.8 (9.9) a ab b a a b a FR 123.4 (181.1) 112.1 (171.2) 0.9 (3.2) 10.4 (25.7) 121.8 (152.5) 85.3 (112) 36.5 (67.3)

150000 AV 18.8 (39.5)a 13.3 (30.1)a 3.8 (11.3)a 1.7 (5.4)ab 46.4 (39.0)a 26.0 (33.7)a 20.4 (20.0)a b b a b b b a c./ml RM 379.1 (232.1) 371.5 (233.2) 1.1 (1.8) 6.5 (12.1) 100.8 (79.5) 80.4 (69.5) 20.3 (31.7) FR 29.4 (59.7)a 28.3 (57.9)a 0.5 (2.0)a 0.5 (2.1)a 253.3 (174.6)c 193.3 (193.9)c 60.0 (66.3)a

37500 AV a a a a a a c./ml RM 13.8 (9.3) 13.7 (9.1) 0.0 (0.0) 0.2 (0.6) 0.0 (0.0) 0.0 (0.0) 0.0 (0.0) a b b b b b FR 17.1 (16.3) 4.8 (13.0) 0.0 (0.0) 12.4 (15.0) 26.9 (39.5) 12.9 (25.4) 14.0 (15.5)

a a a a a a a 24h AV 166.6 (149.5) 144.2 (142.3) 2.7 (5.7) 19.7 (30.0) 19.7 (23.8) 7.6 (18.4) 12.1 (14.8) b b b b b b b RM 6.9 (13.7) 5.8 (12.8) 0.2 (0.4) 0.9 (2.8) 129.8 (76.9) 57.7 (67.6) 72.1 (42.7) c a a a c c a FR 275.7 (213.2) 213.4 (201.6) 15.4 (38.1) 46.9 (61.4) 16.9 (38.0) 0.0 (0.0) 16.9 (38.0)

a a a a a a a 0h AV 3.1 (12.7) 0.7 (3.3) 1.0 (3.3) 1.4 (6.2) 261.5 (153.1) 108.7 (79.5) 152.8 (152.1) b b a a b b b RM 69.6 (78.6) 53.4 (70.6) 1.7 (3.6) 14.5 (33.2) 64.3 (61.9) 43.5 (46.8) 20.8 (32.6) c c b b a ab ab FR 403.4 (308.6) 275.4 (244.7) 17.9 (39.0) 110.1 (109.8) 236.1 (227.0) 122.2 (131.1) 113.9 (156.9) 30 31 Table 5. Analysis of similarity results for the life table parameters of the different Artemia sources, 32 depending on treatment conditions (AV: A. parthenogenetica from Aveiro, RM: A. parthenogenetica 33 from Rio Maior, FR: A. franciscana, c./ml: cells ml1 of Tetraselmis food supply); ANOSIM R and 34 respective p-values are given (—: not enough data for testing). 35 2D NMDS Global AV-RM AV-FR RM-FR Treatment stress R p R P R p R p Base 0.12 0.233 0.001 0.175 0.001 0.311 0.001 0.233 0.001 110ppt 0.10 — — — — — — 0.092 0.092 29ºC 0.16 0.228 0.001 0.199 0.005 0.176 0.004 0.364 0.001 150000 c./ml 0.17 0.400 0.001 0.426 0.001 0.335 0.001 0.476 0.001 37500 c./ml 0.10 — — — — — — 0.456 0.001 24h light 0.14 0.393 0.001 0.530 0.001 0.063 0.026 0.574 0.001 0h light 0.13 0.351 0.001 0.337 0.001 0.424 0.001 0.228 0.001 36

37 Figure(s)

1010°0'W°0'W 5°0'W5°0'W °0'N 45°0'N 45 ±

Aveiro

°0'N Spain 40°0'N 40 Rio Maior

0 50 100 200 Km °0'N 35°0'N 1 35

2 Figure 1. Location of the A. parthenogenetica sampling sites. 3 4 Salinity

AV 70ppt RM 70ppt FR 70ppt

AV 110ppt RM 110ppt FR 110ppt

AV 150ppt RM 150ppt FR 150ppt 0 20 40 60 80 100 120 5 6 Temperature

AV 24 C Series2 RM 24 C Series3 FR 24 C Series4

AV 29 C RM 29 C FR 29 C 0 20 40 60 80 7 8 Food supply (Tetraselmis cells ml1)

AV 300000 RM 300000 FR 300000

AV 150000 Series2 RM 150000 Series3 FR 150000 Series4

AV 37500 RM 37500 FR 37500 0 20 40 60 80 9 10 Light

AV 12h RM 12h FR 12h

AV 24h Series2 RM 24h Series3 FR 24h Series4

AV 0h RM 0h RM 70pptFR 0h 0 20 40 60 80 11 0 20 40 60 80 pre-reproductive reproductive post-reproductive 12 13 14 Figure 2. Mean values (+SD) of pre-reproductive, reproductive and post-reproductive periods, depending 15 on Artemia source and culture conditions; FR: A. franciscana, AV: A. parthenogenetica from Aveiro, 16 RM: A. parthenogenetica from Rio Maior. Notice that there is only 1 observation for AV at 110 ppt. 17 16 10 10 10

8 8 8 12 1 1 1 1   6  6  6 8 4 4 4

broods female broods female broods 2 female broods 2 female broods 2

0 0 0 0 C C C C C C

600 600 600 1000 400 400 400 offspringfemale offspringfemale offspringfemale

offspringfemale 500 200 200 200

0 0 0 0 C C C C C C

AV 0h AV RM 0h RM FR 0h FR AV 12h AV 24h AV RM 12h RM 24h RM FR 12h FR 24h FR AV 24 AV 29 AV RM 24 RM 29 RM FR 24 FR 29 FR AV 70ppt AV RM 70ppt RM FR 70ppt FR FR 150ppt FR AV 37500 AV AV 110ppt AV RM 37500 RM RM 110ppt RM FR 37500 FR FR 110ppt FR AV 150ppt AV RM 150ppt RM AV 300000 AV 150000 AV RM 300000 RM 150000 RM FR 300000 FR 150000 FR 19 8 8 8 8

6 6 6 6

4 4 4 4

2 2 2 2

time between broods (days) betweenbroods time 0 (days) betweenbroods time 0 (days) betweenbroods time 0 (days) betweenbroods time 0 C C C C C C

80 80 80 80 offspringbrood offspringbrood offspringbrood

offspringbrood 40 40 40 40

0 0 0 0 C C C C C C

AV 0h AV RM 0h RM FR 0h FR AV 12h AV 24h AV RM 12h RM 24h RM FR 12h FR 24h FR AV 24 AV 29 AV RM 24 RM 29 RM FR 24 FR 29 FR AV 70ppt AV RM 70ppt RM FR 70ppt FR FR 150ppt FR AV 110ppt AV RM 110ppt RM FR 110ppt FR AV 150ppt AV 37500 AV RM 150ppt RM RM 37500 RM FR 37500 FR AV 300000 AV 150000 AV RM 300000 RM 150000 RM FR 300000 FR 150000 FR 21 22 Figure 3. Mean values and SD of brood and offspring parameters depending on Artemia source and 23 culture conditions; AV: A. parthenogenetica from Aveiro, RM: A. parthenogenetica from Rio Maior, FR: 24 A. franciscana. The first set of bars in each plot refers to the base treatment condition. Only females 25 producing at least three broods were considered for the calculation of the time between broods. 26 100% 100%

80% 80%

60% 60% badcysMN badcysMN goodcysMN goodcysMN 40% 40% abortembMN abortembMN bad cysts deadembMN deadembMN good cysts lifeembMN 20% 20%lifeembMN abortive embryos dead embryos life embryos 0% 0% C C C C C C RM24 RM29 AV 24 AV 24 FR 29 AV 29 FR RM 70ppt RM AV 70ppt AV FR 70ppt FR RM 110ppt RM 150ppt RM 27 110ppt AV 110ppt FR 150ppt AV 150ppt FR 100% 100%

80% 80%

60% 60% badcysMN badcysMN goodcysMN 40% 40%goodcysMN abortembMN abortembMN deadembMN deadembMN 20% 20% lifeembMN lifeembMN

0% 0% RM 0h RM AV 0h AV 0h FR RM 12h RM 24h RM AV 12h AV 12h FR 24h AV 24h FR RM 37500 RM AV 37500 AV 37500 FR RM 300000 RM 150000 RM AV 300000 AV 300000 FR 150000 AV 150000 FR

28 29 Figure 4. Mean offspring composition, depending on Artemia source and culture conditions; AV: 30 A. parthenogenetica from Aveiro, RM: A. parthenogenetica from Rio Maior, FR: A. franciscana. Only 31 the offspring from females producing at least three broods was considered. 32 33 34 2.0 ovoviviparous 70ppt life emb. 24°C 300000 cells ml1 12 h light AV PoRP PrRP RM FR LS abort. emb. dead emb. offspr./brood RP broods female .5% offspr./female 19

RM bad cysts good cysts FR 35 -1.0 oviparous 40.8% -1.0 2.5 1.5 ovoviviparous 100ppt life emb. abort. emb.

offspr./brood dead emb.

PoRP offspr./female %

23.7 broods/female RP PrRP LS good cysts AV RM bad cysts oviparous 36 FR

-1.3 42.1%

1.5 -1.5 PrRP 29C2.0 LS PoRP dead emb.

life emb. ovoviviparous broods/female RP

20.9 offspr./female abort. emb. offspr./brood bad cysts

AV oviparous good cysts 37 RM 46.7% FR 38 -1.5 39 Figure -1.0 5. First two axes ofthe correlation-based PCAs of Artemia2.5 life table parameters, depending on 40 treatment conditions. Samples are identified according to the Artemia source (AV: A. parthenogenetica 41 from Aveiro, RM: A. parthenogenetica from Rio Maior, FR: A. franciscana, emb.: embryos, abort.: 42 abortive) and percentages of variance explained by the axes are given. 43 1.8 150000 cells ml1 oviparous good cysts

bad cysts offspr./brood

RP offspr./female % broods/female 16.9 LS PrRP dead emb. PoRP

abort. emb. AV ovoviparous 44 life emb. RM 45.5% FR

1.8 1 -1.2 37500 cells ml ovoviparous -1.0 PoRP 1.8 abort.emb. offspr./brood LS

life emb. offspr./female

RP % broods/female bad cysts

22.1 oviparous PrRP good cysts

45 48.6% RM oviparous 24 h light FR -1.5 1.5 -1.0bad cysts good cysts 3.0

PrRP offspr./brood LS RP offspr./female broods/female

% life emb. 20.0 dead emb. ovoviviparous

abort. emb. PoRP AV 46 RM 38.6% FR 47 -1.2 48 Figure 5.-1.0 (Continuation). 2.5 49 2.0

0 h light ovoviviparous abort. emb. life emb. dead emb. offspr./brood

% offspr./female 20.5 PrRP broods/female RP PoRP good cysts LS

AV bad cysts oviparous RM 50 47.2% FR 51 -1.5 52 Figure 5.-1.0 (Continuation). 2.5 53