The life cycle of the alien boatman Trichocorixa verticalis (Hemiptera, Corixidae) in saline and hypersaline wetlands of south-west Spain
Trichocorixa verticalis (Corixidae) is native to North America but is well established as an alien in the Western Mediterranean region, where it is invasive in permanent coastal wetlands with high salinities. We investigated the annual cycle and generation time of T. verticalis in the introduced range in south-west Spain, through a combination of field surveys and laboratory experiments. Field surveys were conducted on a monthly basis over 1 year in three saline fish ponds in Doñana and four hypersaline salt ponds in the Odiel marshes. Adults were present all year round, whereas nymphs were only absent in August, when temperatures and salinities were high. Adult sex ratios were idiosyncratic and often male or female biased for a given location and month. Adults were smaller during summer months. Laboratory experiments revealed an oviposition rate of 11.5 eggs per day and a generation time of about 54 days from egg to adult, suggesting T. verticalis may complete around six generations per year in permanent wetlands. A combination of a high oviposition rate and continuous reproduction throughout the year gives T. verticalis an advantage over native corixid competitors (Sigara spp), and appears to explain the success of this alien aquatic insect.
KeywordsTrichocorixa verticalis Fecundity Fish ponds Generation time Invasive species Salt ponds Life cycle
A major question in invasion biology is what factors determine whether a species becomes invasive or not (Drake et al., 1989). Those factors include both the species traits and aspects of the habitat to which the species is introduced (Richardson & Pyšek, 2006). Thus, postulated mechanisms of invasions have focused either on (a) traits of the invaders themselves (e.g., the “tens rule”) and the concept of species invasiveness (Williamson & Fitter, 1996); or (b) the ecological interactions between the invader and the recipient community (e.g., “invasional meltdown,” “novel weapons,” “enemy release,” hypotheses, see Jeschke et al., 2012 for review) and the invasibility of an ecosystem. Life history traits are of key importance in determining the invasiveness of a species (Rajagopal et al., 1999; Grabowski et al., 2007). Short generation time, early sexual maturity, high fecundity, large body size, and euryhalinity, among others can enable the establishment and population increase of an alien species in new aquatic environments, allowing one species to outcompete another (Bij de Vaate et al., 2002; Grabowski et al., 2007).
The Corixidae is the largest family of aquatic Hemiptera, with species inhabiting different habitats, such as lotic or lentic, continental or coastal, at different salinities, from freshwater to hypersaline waters. They constitute an important functional element in such ecosystems, contributing significantly to energy flow through the ecosystem. Corixids are detritivores responsible for the cycling of the organic matter (Kumari & Kumer, 2003), and are also key predators of zooplankton at intermediate levels in the food web with the potential for cascading effects (Henrikson & Oscarson, 1981; Simonis, 2013); in turn, they are an important part of the diet of other invertebrates, fish, and waterbirds (Henrikson & Oscarson, 1978; Giles et al., 1990; Euliss & Jarvis, 1991).
Trichocorixa verticalis (Fieber, 1851) is a small euryhaline corixid (Hemiptera) (length 3.5–5 mm) originally distributed across a wide latitudinal range in North America and the Caribbean, but the subspecies T. verticalis verticalis is an alien in South Africa, New Caledonia, and the Western Mediterranean (Morocco, Spain and Portugal; Guareschi et al., 2013). In its native range, Trichocorixa verticalis occurs in coastal and inland saline habitats up to 70–80 g l−1 (Hutchinson, 1959; Wurtsbaugh, 1992; Jansson, 2002). The life history of this species (e.g., number of generations per year, overwintering strategy) can be expected to vary considerably over such a broad natural range (covering c.25 degrees of latitude, Guareschi et al., 2013). Some studies have reported two or three generations per year (Tones, 1977; Kelts, 1979), and in Canadian lakes that freeze in winter, T. verticalis interiores produces eggs that go through diapause in winter and early spring (Tones, 1977). Within the Trichocorixa genus, T. kanza and T. calva overwinter as adults (Sailer, 1948).
In Europe, T. verticalis verticalis is now well established as an alien in the southwestern part of the Iberian Peninsula, but is projected to expand widely in coastal Europe over time (Guareschi et al., 2013). In south-west Spain, it is found at salinities ranging from 0.6 to 75 g l−1 (Rodríguez-Pérez et al., 2009; Coccia et al., 2016b; Carbonell et al., 2017), and has become invasive in saline and hypersaline permanent water bodies, where it outcompetes native halotolerant species such as Sigara selecta or S. stagnalis. In salt ponds, T. verticalis is an effective predator of brine shrimp Artemia (Céspedes et al., 2017). In contrast, in temporary ponds of lower salinity, T. verticalis is less abundant, and co-occurs with native Sigara (mainly S. scripta and S. lateralis) along different parts of the salinity gradient (Rodríguez-Pérez et al., 2009; Carbonell et al., 2016, 2017), with overlap in their dietary niches (Coccia et al., 2016a).
There remains a lack of basic studies on the ecology of T. verticalis. As a result, the mechanisms underlying the success and invasiveness of this alien insect, and the reasons why permanent, saline habitats are more invasible, are currently unclear. T. verticalis is smaller than Sigara spp. and unable to displace them through aggressive behavior, and is more susceptible to size-limited predators such as Odonata larvae (Coccia et al., 2014; Carbonell et al., 2017). Although it coincides with other invasive species in the introduced range (Rodríguez-Pérez & Green, 2012; Walton et al., 2015), there is no evidence that these species facilitate invasion by T. verticalis as would be suggested by the invasional meltdown hypothesis (Simberloff & Von Holle, 1999). For example, alien fish predators showed no preference between T. verticalis and native S. lateralis (Coccia et al., 2014). Similarly, there is no evidence to support the “Enemy release hypothesis” (see Jeschke et al., 2012), since T. verticalis is more susceptible to ectoparasitic water mites than are native Corixidae, although these parasites are absent in saline and hypersaline habitats (Sánchez et al., 2015). T. verticalis shows high plasticity in its thermal ecology and is better able to acclimate to high salinities and high temperatures than S. lateralis (Coccia et al., 2013). Its eggs are more tolerant to high salinities than co-occurring Sigara species, and its wing morphology suggests that it has a stronger dispersal ability (Carbonell et al., 2016).
Based on preliminary data from different seasons (Rodríguez-Pérez & Green, 2012) and on laboratory studies by Carbonell et al., 2016, we hypothesized that the invasiveness of T. verticalis in permanent, saline wetlands could be related to higher rates of population growth than native Sigara species, because it is able to reproduce throughout the annual cycle and/or has a higher fecundity. In order to test this hypothesis, herein we investigate the life cycle of T. verticalis, and its population ecology during one annual cycle in permanent hypersaline salt ponds and saline fish ponds in south-west Spain. Our specific objectives were (i) to quantify its life cycle (fecundity and duration of each developmental instar) in the laboratory, (ii) to assess spatial and seasonal variation in population density and reproductive activity in a selection of ponds representative of habitats where the species is invasive, and relate this variation with environmental variables, (iii) to quantify seasonal variation in adult sex ratio and body size.
Study areas in south-west Spain
The Odiel Marshes (37°17′N 06°55′W, Fig. 1) are located in Huelva province at the mouths of the rivers Odiel and Tinto. These marshes are protected as a Biosphere Reserve, Ramsar site, Natura 2000 site, and Natural Park owing to their importance for migratory waterbirds (Sánchez et al., 2006a, b). They contain 7,185 ha of intertidal mudflats, of which 1,120 ha have been transformed into industrial salt ponds. Seawater is pumped through a series of ponds and salinity increases via evaporation until crystallization. Water first circulates through a series of primary evaporation ponds (salinity 25–70 g l−1) where corixids are abundant, followed by secondary evaporation ponds (salinity 40–125 g l−1) where brine shrimp Artemia are abundant but corixids are limited to the ponds of lower salinity (Sánchez et al., 2006a). Finally, there is a series of crystallization ponds where salt precipitates and corixids are absent. T. verticalis is the only corixid species recorded at this site, and the date of invasion is unknown. The invertebrate community was previously studied by Sánchez et al. (2006a). We studied four salt ponds with different depths and salinity ranges (E1 = 50–60 cm and 35–65 g l−1; E2 = 10–15 cm and 39–70 g l−1; E3 = 10–40 cm and 40–75 g l−1; E4 = 15–25 cm and 60–85 g l−1; see Fig. 1 and Table S1 for more details). Submerged vegetation was dominated by the macrophytes Ruppia cirrhosa and Althenia filiformis, and the algae Chaetomorpha and Enteromorpha. Shorelines were dominated by the halophytes Arthrocnemum macrostachyum and Salicornia ramosissima (Chenopodiaceae). As at VLP, the waterbirds at Odiel are likely to feed partly on Corixidae (Fuentes et al., 2004; Boros et al., 2006).
In VLP, T. verticalis were collected in three ponds from February 2011 to January 2012 inclusive (hereafter 2012), whereas in Odiel T. verticalis were collected in four ponds from November 2014 to October 2015 (hereafter 2015). Samples were collected once a month using a D-framed pond net (500 mm mesh; 0.16 m2 core) while wading. We applied a semi-qualitative sampling method by sweeping a hand-net in two points along the shorelines (i.e., two replicate samples in each pond). At each of these points, a series of sweeps were carried out over a 1 m transect for a total of 30 s. We collected samples between 10:00 and 14:00 h in a fixed order (ponds G3, A3, A7 at VLP and E4, E3, E1 and E2 at Odiel, see Fig. 1).
After collection, T. verticalis samples were transported to the laboratory inside plastic containers filled with water and submerged vegetation from the collection site. At the laboratory, water was substituted with 70% of ethanol for sample preservation until analysis. No other corixid species were recorded in Odiel samples. In VLP samples, small numbers of Sigara spp. were recorded in ponds A7 and G3 in December only, but were not quantified. When sampling, we measured in situ temperature (with a WTW 340i multiprobe) and salinity (with a refractometer: RHS-28) and we collected water samples for analysis of concentrations of chlorophyll a (as a measure of phytoplankton abundance) and nutrients (N and P). Once at the laboratory, water samples were analyzed as follows (see results in Table S1): Chlorophyll a concentration (μg l−1) was determined using methanol extraction (Talling & Driver, 1963), total nitrogen concentration (Total N) was measured by digestion with potassium persulfate (Sims et al., 1995), and total phosphorus concentration (Total P) by the phosphomolybdate method (APHA, 1995). Dissolved nutrient concentrations (nitrates, nitrites, ammonium, and phosphate) were measured after water filtration through a standard glass filter, using a SEAL Analytical Auto Analyzer 3HR and a SYSTEA Micromac-1000. The filtered samples were frozen at − 20°C for several weeks prior to analysis. Daily air temperature data were extracted from Doñana’s ICTS (Singular Scientific-Technical Infrastructure) repository held by the Estación Biológica de Doñana (EBD-CSIC).
Sex ratios were quantified for T. verticalis adults collected from two ponds (G3 and A7) in VLP and all ponds in Odiel. While all adults were sexed in samples with less than 50 individuals, for larger samples, we first placed the preserved adults in a large petri dish and then thoroughly spread them around in a random fashion. Then, we divided the petri disc into four quadrants and selected one quadrant as a subsample. From this quadrant, we randomly picked out 50 individuals for sexing.
In samples from A7 pond in VLP, 20 adult individuals (10 from each sex, when possible) were measured per month for total body length (from the front of the head until the apex of abdomen) using an image analysis system (Axio Vision 4.8.). In samples from A7 and all Odiel ponds, nymphs were identified to instars following Melo & Scheibler (2011). When nymphal numbers were high in samples from A7 (i.e., > 150), a subsample of 110 individuals was randomly selected in the above manner and then assigned to instars. The proportions for each instar were then extrapolated to the total sample to estimate the total numbers of nymphs from each instar.
Experimental study of life histories
Indoor microcosms to quantify the duration of larval instars
Adults of T. verticalis were collected on April 9, 2013 in VLP pond A7 and then transported to the laboratory as above. There, 20 individuals (10 males and 10 females) were placed in each of three 3 L aquaria filled with 10 g l−1 solution created from deionized water mixed with salts (Ocean Fish, Prodac®. Citadella Pd, Italy). Aquaria were kept in a controlled climate chamber set at 22 ± 1°C. In each aquarium, one stone and two 10 × 10 cm pieces of rigid plastic (1 cm mesh) were placed to facilitate egg laying. Aquaria were covered with soft 1 mm mesh to prevent corixids from escaping. Every two days, 1 ml of water containing dissolved algae (Tetraselmis chuii—Easy Algae®) was added to each aquaria with a pipette, and frozen chironomid larvae were provided ad libitum to feed nymphs and adults, respectively. The water level was maintained by adding deionized water when necessary, and water in the aquaria was completely replaced once a week after sieving out nymphs and adults. Each day, we checked for the presence of eggs. We also counted all live and dead adults to assess mortality, and the number of different nymph instars to quantify the time taken for development of each instar. Once the first nymphs reached adulthood, the experiment was ended for that aquarium.
Outdoor microcosms to quantify fecundity and the time taken for adults to mature
We conducted a second experiment aimed at quantifying the time interval between reaching adulthood, mating and egg laying, and to quantify fecundity. Given that corixid mortality was high in our indoor experiment, perhaps because artificial natural light may have affected periphyton growth (i.e., food supply), we decided to conduct this second one outdoors. On May 17, 2016, we collected T. verticalis nymphs in VLP pond A7, together with sediments and 50 l of pond water (salinity 22 g l−1). Once in the laboratory, nymphs of the IVth and Vth instars were placed in groups of 10 nymphs per 500 ml microcosm (height 103 × diameter 91 mm). Each microcosm was provided with 5 cm of sediments, filled with 350 ml of saline pond water, and covered with a fine mesh to prevent escape when adulthood was reached. Approximately, 900 nymphs were placed in a total of 90 microcosms maintained on the roof of the Doñana Biological Station (EBD-CSIC). This outdoor experiment was started on May 18, 2016. We measured temperature continuously with two data loggers (from 18th May to 11th June, mean daily Tª max = 43.92 ± 0.85 (± s.e.); Tª min = 16.04 ± 0.46; Tª mean = 23.19 ± 0.46). Microcosms spent part of the day in direct sunlight, and nymphs often buried themselves in the sediments where peak temperatures may have been lower. Daily, we checked the nymphs and added 1 ml of algae (Tetraselmis chuii—Easy Algae®) as food. Frozen chironomid larvae were added every 2 days. Water was replaced frequently.
When nymphs reached adulthood, we first checked their sex, and then one female + one male were placed together in a new microcosm with identical environmental characteristics. In total, 26 pairs were created. Each microcosm contained a wooden stick for egg laying. Daily, we quantified adult survival and monitored mating and egg laying. When an adult died, it was replaced by another individual of the same sex. Egg laying was monitored for each female until it died. Once the first eggs were laid, more eggs were laid on successive days or after an interval of up to 4 days. The experiment was finished on 11th June, after none of the pairs was observed to have laid an egg or form copulas during the previous 6 days.
We carried out correlation matrices between environmental variables at the time of monthly sampling and the abundances of adults and nymphs in each pond (Tables S1, S5) as well as the body length of adults (female and males) (Fig. 5 and Table S6), using non-parametric Spearman correlations in the package corrplot R-studio (ver. 1.1.453 – © 2009-2018 RStudio, Inc.).
Confidence intervals for adult sex ratios were calculated at http://www.easycalculation.com. To analyze temporal variation in adult body size and to test for differences between sexes, we used an ANOVA with the total length of adults as the dependent variable and month, sex, and their interaction as explanatory variables. Significant differences between sexes and months were determined by post hoc analysis employing Bonferroni correction. Analyses were performed using Statistica (software version 12, IBM, StatSoft CR; http://www.statsoft.cz/).
To visualize the relative abundance of adults and nymphs throughout the annual cycle, we constructed monthly histograms of instar-frequency distributions using the data from field samples. Separate histograms for VLP and Odiel were constructed, using the FISAT program (FAO-ICLARM, stock assessment tools, ver. 1.01).
Density of T. verticalis adults and nymphs throughout the annual cycle
In VLP (Fig. 2A), adult abundance remained relatively stable from August to October, whereas nymphs practically disappeared. Nymph density peaked from February to May. In Odiel (Fig. 2B), both nymphs and adults were most abundant from May to July, then again in November and December, these being periods when salinity was lower (about 50 g l−1) than during months inbetween. Although December had relatively low temperatures, the mean daily maximum air temperature was about 17°C.
Relationship between density, salinity, and other environmental variables
When relating abundance in each of the seven study ponds to environmental variation over the annual cycle, we found more evidence for limiting effects of high temperatures and salinities at Odiel than at VLP (Table S5). At VLP, the only significant relationship between corixid abundance and air temperature was a relatively weak negative correlation between nymph abundance and minimum temperature at A7). There were negative correlations between abundance and salinity for one pond each for nymphs (A7) and adults (A3). At Odiel, negative correlations between salinity and the abundance of adults and nymphs were recorded at ponds E3 and E4 and were particularly strong in E2. In contrast, at E1, abundance of adults and nymphs had a strong negative correlation with air temperature, a result also recorded at E4 and for adults at E3 (Table S5)).
Within a given pond, correlations between corixid abundance and parameters indicating resource availability such as chlorophyll a, Total P, and Total N were inconsistent and often negative (Table S5), suggesting that such correlations were largely driven by confounding effects of temperature and evaporation. Chlorophyll a was significantly correlated with adult abundance in three ponds (two positive, one negative) and with nymph abundance in three ponds (one positive, two negative).
Adult sex ratio
Seasonal variation in adult length
Seasonal changes in abundance of life history stages
Duration of life history stages in laboratory conditions
We have shown T. verticalis to be abundant throughout the annual cycle in permanent, saline, and hypersaline wetlands in south-west Spain, and found evidence that its success as an invader is explained by an ability to reproduce throughout the year, and a high fecundity. Our study sites typify the habitats in which T. verticalis is invasive in the Western Mediterranean region. We were not able to compare alien and native corixid species in the same wetlands because native Sigara species were almost absent from our study area.
Continuous reproduction, voltinism, and generation time
To date, T. verticalis is the only corixid species in the Western Mediterranean observed to have nymphs in winter months. The abundance of both nymphs and adults almost all year long suggests that near-continuous reproduction allows T. verticalis to outcompete native Sigara species (S. lateralis, S. selecta, S. scripta, and S. stagnalis) in permanent habitats by providing a high population growth rate. In contrast, these Sigara species do not breed through the winter, and they overwinter as adults (Perán, 1997; Barahona et al., 2005). Egg deposition by corixids in temperate zones usually occurs in the spring (univoltine species) and summer (bivoltine species) (Griffith, 1945; Fernando, 1959). Even when polivoltinism is present, corixids typically overwinter as adults. In contrast, the abundance of T. verticalis nymphs of all instars throughout the winter in our study sites (Fig. 6) can only be explained by winter reproduction.
The breeding phenology and life cycles of corixids can be expected to vary considerably between different populations of species such as T. verticalis that have an extensive native range. Intraspecific flexibility in the number of generations per year is likely in response to thermal differences between habitats at different latitudes or altitudes (Ward & Stanford, 1982). The ability of T. verticalis to breed through winter can largely explain its invasiveness in permanent, saline wetlands in coastal Iberia where it is not interrupted by desiccation and where populations can accumulate through successive generations. This is likely to be related to the mild winter temperatures in this part of Spain where water temperatures remain well above freezing point and winter temperatures have increased in recent decades (Espinar et al., 2015). T. verticalis is exposed to much lower winter temperatures in its native range, and hence may be preadapted to breeding during the Iberian winter. Indeed, we have found evidence that reproduction may be halted in high summer, or at least be inhibited by high temperatures and not low ones (Figs. 2 and Table S5). However, reproduction throughout the winter is unlikely if the species expands to northern Europe (as predicted by current models, Guareschi et al., 2013) where winters are much colder.
Our laboratory experiments show that a generation takes about 54 days (under conditions of 10–20 g l−1 salinity; and mean water temperature of 21–23°C), suggesting that there could be time to complete at least six generations per year in the field under continuous reproduction and with overlapping cohorts. The available literature suggests that its competitors S. scripta and S. selecta have 4–5 generations per year in the Western Mediterranean (Perán, 1997; Barahona et al., 2005). In hyposaline and mesosaline Mediterranean streams in eastern Spain, S. scripta and S. selecta had multivoltine cycles and four overlapping cohorts, each with a generation time of 2–3 months (Perán, 1997; Barahona et al., 2005). Nymphs of S. selecta occurred in almost all instars for most of the year, except during winter, when only adults were recorded. The mean temperature threshold for mating and oviposition in S. selecta was about 13°C.
We measured the duration of each developmental instar of T. verticalis under a single temperature regime in microcosms, to provide baseline information on the development time for this alien species. Development times are likely to be highly dependent on food supply and temperature, thus we must be cautious when extrapolating our experimental results from microcosms to field conditions. It is likely that generation time in our study wetlands is longer in winter when temperatures are relatively low, and this would be consistent with seasonal changes recorded in body length (see below). Sexual development, egg laying, and metabolic rates were highly dependent on temperature in Sigara alternata (Sweeney & Schnack, 1977). While keeping several T. verticalis individuals together in the same laboratory mesocosm during our indoor experiment, we only recorded the first observation for each development instar, because it was not possible to distinguish between individuals. Hence, in some cases, our data may underestimate the development times of average individuals.
The high oviposition rate we recorded for T. verticalis (11.5 eggs/day per female) supports earlier findings from the laboratory experiment of Carbonell et al. (2016), who found that T. verticalis showed at least twice the laying rate recorded for S. lateralis, S. scripta, and S. selecta. This marked difference in fecundity between the alien and Sigara spp. was retained when correcting for hatching success (Carbonell et al., 2016), which we were unable to quantify in our own study.
When considering days from the laying of the first to the last egg for each individual (≥ 1 egg laid), we found that female T. verticalis laid 7–24.8 eggs per day (mean 11.52 + 0.8 s.e.) at 20 g l−1, and comparable data from Carbonell et al., 2016 reveal that female T. verticalis laid 1.5–53 (mean 13.35 + 5.64 s.e.) eggs per day at 25 g l−1. Furthermore, in an unpublished experiment conducted more recently in our lab at a salinity of 15 g l−1 and a temperature of 20°C, female S. lateralis laid only 0.25–4 (mean 1.7 + 0.81 s.e.) eggs per day (data from Céspedes et al., submitted), showing that T. verticalis are consistently more fecund than their competitors, at least at high salinities. The exceptional fecundity and high aerial dispersal ability (Carbonell et al., 2016) of T. verticalis are also likely to promote its rapid expansion along coastal areas of the Iberian Peninsula and beyond (Guareschi et al., 2013).
Our observations suggest that T. verticalis females only have one bout of egg laying which lasts no more than four days. However, under some field conditions, or at some times of the year, females may live longer and lay eggs over an extended period, especially as our experiment was done outdoors when microcosms were exposed to hot maximum temperatures. In a related laboratory study on S. selecta at 18–22°C and 37 g l−1, the mean oviposition period was about 3 weeks, although some females extended this to more than 1 month. During this period, days with intense oviposition alternated with periods of little or no oviposition. Mean total fecundity was 31 eggs and the mean rate of oviposition was 1.4 eggs per day (Barahona et al., 2005).
Adult sex ratio and body length
Adult T. verticalis sex ratio was often female biased in Odiel, but often male biased in VLP. Seasonal patterns were inconsistent, and could have multiple causes related to the relative survival rates of males and females at different instars, sex ratio at hatching, or different dispersal rates between sexes (Boda & Csabai, 2009; Carbonell et al., 2016). Since they are smaller, males may be subject to different predation pressures, although there is no evidence for this from a laboratory experiment with fish predation (Coccia et al., 2014). In a previous study of T. verticalis in native Canada in lakes that freeze in winter (i.e., very different temperatures to our study area), biased sex ratios were recorded from May to July because females reached sexual maturity a week ahead of males (Aiken & Malatestinic, 1995). Barahona et al. (2005) found the sex ratio in S. selecta to vary, being female biased for most of the year but male biased or balanced in spring, and suggested that this was due to the effect of higher female longevity following the emergence of the first spring generation.
We found female T. verticalis to be consistently larger than males, as expected (Coccia et al., 2013). However, we also found that adult body size varied between months, and was lower during summer. Development is likely to be more rapid during warmer months, leading to maturity at a smaller size when the benefits of early maturity may be greater e.g., to enable rapid dispersal to colonize other waterbodies.
Environmental conditions that favor or limit the reproduction of T. verticalis
Generally, the densities of T. verticalis recorded in saline VLP and hypersaline Odiel were similar (Fig. 3), showing T. verticalis thrives over an extensive salinity range of 8–74 g l−1. There is very limited reproduction from August to October when temperatures and salinities are both high (Fig. 2, Table S5). In VLP, reproduction is most pronounced in spring when temperatures and salinities begin to increase. In hypersaline Odiel, reproduction peaks in November when winter rainfall reduces salinity. Differences between our seven study ponds in overall density of adults or nymphs along the annual cycle were not explained by salinity (Fig. 3). Nevertheless, changes in corixid abundance between months within an individual pond were often negatively correlated with salinity, especially in hypersaline ponds, suggesting that extreme salinities can limit abundance and reproduction, although there was also evidence that high temperatures can be more limiting (Table S5). There was no consistent evidence from nutrient or chlorophyll concentrations to suggest that T. verticalis abundance was determined by pond productivity.
In the native range, T. verticalis is reportedly unable to reproduce in abundance at salinities of > 60 g l−1 (Wurtsbaugh, 1992; Simonis, 2013). This is roughly consistent with our results, including the finding that the ratio of nymphs to adults dropped in our most saline pond E4, above 70 g l−1 (Fig. 3). T. verticalis is rarely recorded in salt ponds of salinities > 80 g.l−1 at Odiel (Sánchez et al., 2006a). Depth is also important, and in Mediterranean wetlands corixid densities are generally higher in shallow areas of below 30 cm (Munoz-Fuentes et al. 2013). T. verticalis density is low at depths of over 50 cm, and this may explain why Odiel pond E1 had a low density, since it was unusually deep (see methods). Preference for shallow areas may help to explain why T. verticalis densities drop off at the hottest times of the year, since water temperatures in the shallows then often exceed air temperatures.
Habitat temporality, ectoparasites, and relationship with Artemia
Temporary habitats are abundant within and around Doñana in southwestern Spain (Green et al., 2018). In these habitats, T. verticalis is rarely dominant and coexists with abundant S. lateralis and S. scripta (Coccia et al., 2016a, b; Carbonell et al., 2017). The lower success of T. verticalis in temporary ponds is likely to be related to the rebooting of competition each year after ponds reflood, since all corixid species then need to recolonize the sites and T. verticalis retains no numerical advantage from higher fecundity in previous years. Furthermore, these habitats have a lower salinity, and are frequented by ectoparasitic water mites that are particularly likely to infest the alien species (Sánchez et al., 2015).
In salt ponds, T. verticalis preys on Artemia (Céspedes et al., 2017), and in Iberian salt ponds there is an ongoing replacement of native Artemia by the highly invasive American A. franciscana (Horváth et al., 2018), which is syntopic with T. verticalis in their native range. There is a strong parallel between the invasions by American corixids and brine shrimps, since A. franciscana continues to reproduce throughout the winter in Iberian salt ponds unlike the native Artemia, and is also more fecund than them (Redon et al., 2015). The invasion of A. franciscana might potentially benefit T. verticalis by boosting their food supply during winter, but there is no evidence that they facilitated the invasion by the corixid (e.g., Artemia are absent from VLP).
Our study has improved our understanding of the life cycle of T. verticalis within the introduced range. The invasion success of T. verticalis in permanent, saline wetlands is due to a particularly high population growth rate that allows the alien to dominate the corixid community. There was no strong trend in abundance and reproduction across the salinity range from 8 to 74 g l−1, except that little or no reproduction was recorded at over 70 g l−1. Reproduction was confirmed throughout winter months, when average daily mean air temperatures were around 9.4°C and average daily maxima were around 16.2°C, and no such winter reproduction is reported for native Corixidae. Months of lowest abundance are those of highest salinities and/or of highest temperatures. Laboratory experiments and field sampling suggest that T. verticalis could complete one or two more generations per year than competing Sigara species. T. verticalis also has higher fecundity than native Sigara spp. Future studies should compare the life cycle of T. verticalis and native Sigara spp in temporary habitats of lower salinity.
E. Martinez (Director of Marismas del Odiel Natural Park) and Doñana Natural Space provided permission for fieldwork. J. Miguel Medialdea and Pesquerías Isla Mayor, S.A. provided facilities in VLP. Miguel Lozano Terol, Raquel López Luque, Natalia Ospina-Alvarez, and Simona Kacmarcikova helped with laboratory and fieldwork. Andres Millán and Josefa Velasco provided helpful advice. Ruben Izquierdo and the “MiBuho” company provided the graphic design for Fig. 7. The staff of the Aquatic Ecology (LEA-EBD) and GIS and Remote Sensing (LAST-EBD) laboratories of EBD-CSIC provided essential support. This research was funded by the Consejería de Innovación, Ciencia y Empresa, Junta de Andalucía project (P10-RNM-6262) to AJG, a Severo Ochoa predoctoral contract (SVP-2013-067595) from the Spanish Ministry of Science and Innovation (MICINN) to VC, a JAE predoctoral grant from CSIC and a post doc project 3160330 financed by FONDECYT to CC, a predoctoral FPU grant to JAC and a Ramón y Cajal postdoctoral contract from MICINN to MIS. Two anonymous referees greatly improved an earlier version of the manuscript.
- Apha, A., 1995. WPCF, standard methods for the examination of water and wastewater. American Public Health Association, Washington, DC.Google Scholar
- Céspedes, V., M. I. Sánchez & A. J. Green, 2017. Predator-prey interactions between native brine shrimp Artemia parthenogenetica and the alien boatman Trichocorixa verticalis: influence of salinity, predator sex, and size, abundance and parasitic status of prey. PeerJ 5: e3554.CrossRefGoogle Scholar
- Céspedes, V., Valdecasas, A.G., Green, A.J. & Sánchez, M.I. submitted. Effects of water mite parasites and salinity stress on aquatic insects.Google Scholar
- Coccia, C., P. Calosi, L. Boyero, A. J. Green & D. T. Bilton, 2013. Does ecophysiology determine invasion success? A comparison between the invasive boatman Trichocorixa verticalis verticalis and the native Sigara lateralis (Hemiptera, Corixidae) in South-West Spain. PLoS ONE 8(5): e63105.CrossRefGoogle Scholar
- Drake, J. A., H. A. Mooney, F. di Castri, R. H. Groves, F. J. Kruger, M. Rejmánek & M. Williamson (eds), 1989. Biological invasions: a global perspective. Wiley, Chichester.Google Scholar
- Fernando, C. H., 1959. The colonization of small freshwater habitats by aquatic insects, 2. Hemiptera (The water bugs). Ceylon Journal of Science 2: 5–32.Google Scholar
- Fuentes, C., M. I. Sanchez, N. Selva & A. J. Green, 2004. The diet of the Marbled Teal Marmaronetta angustirostris in southern Alicante, eastern Spain. Revue D’ Ecologie-La Terre Et La Vie 59: 475–490.Google Scholar
- Giles, N., M. Street & R. M. Wright, 1990. Diet composition and prey preference of tench, Tinca tinca (L.), common bream, Abramis brama (L.), perch, Perca fluviatilis L. and roach, Rutilus rutilus (L.), in two contrasting gravel pit lakes: potential trophic overlap with wildfowl. Journal of Fish Biology 37(6): 945–957.CrossRefGoogle Scholar
- Green, A. J., J. Bustamante, G. F. E. Janss, R. Fernández-Zamudio & C. Díaz-Paniagua, 2018. Doñana Wetlands (Spain). In Finlayson, C. M., G. R. Milton, R. C. Prentice & N. C. Davidson (eds), The Wetland Book: II: Distribution, Description and Conservation. Springer, New York. https://doi.org/10.1007/978-94-007-4001-3_139.CrossRefGoogle Scholar
- Griffith, M. E., 1945. The environment, life history and structure of the water boatmen Ramphocorixa acuminata (Uhler) (Hemiptera Corixidae). The University of Kansas Science Bulletin 30: 241–365.Google Scholar
- Guareschi, S., C. Coccia, D. Sánchez-Fernández, J. A. Carbonell, J. Velasco, L. Boyero, A. J. Green & A. Millán, 2013. How far could the alien boatman Trichocorixa verticalis verticalis spread? Worldwide estimation of its current and future potential distribution. PLoS ONE 8: e59757.CrossRefGoogle Scholar
- Henrikson, L. & H. G. Oscarson, 1981. Corixids (Hemiptera—Heteroptera), the new top predators in acidified lakes: with 2 tables in the text. Internationale Vereinigung für theoretische und angewandte Limnologie: Verhandlungen 21(3): 1616–1620.Google Scholar
- Jansson, A., 2002. New records of Corixidae (Heteroptera) from northeastern USA and eastern Canada, with one new synonymy. Entomologica Fennica 13: 85–88.Google Scholar
- Jeschke, J. M., Gómez Aparicio, L., Haider, S., Heger, T., Lortie, C. J., Pyšek, P., & Strayer, D. L. 2012. Support for major hypotheses in invasion biology is uneven and declining.Google Scholar
- Kumari, P. & A. Kumer, 2003. Biodiversity of aquatic insects of Jharkhand state with reference to their ecological niche. Environment, pollution and management. APH Publishing Corporation, Delhi: 453–460.Google Scholar
- Melo, M. C. & E. E. Scheibler, 2011. Description of the immature instars of Sigara (Tropocorixa) jensenhaarupi (Hemiptera: Heteroptera: Corixidae: Corixini), with ecological notes. Revista Mexicana de Biodiversidad 82(1): 117–130.Google Scholar
- Perán, A., 1997. Ciclos de vida y producción secundaria en un río de características semiáridas (Río Chícamo, SE de la Península Ibérica): El caso de Caenis luctuosa y Sigara scripta.M.S. dissertation, Universidad de Murcia, Murcia.Google Scholar
- Rodríguez-Pérez, H., M. Florencio, C. Gómez-Rodríguez, A. J. Green, C. Díaz-Paniagua & L. Serrano, 2009. Monitoring the invasion of the aquatic bug Trichocorixa verticalis verticalis (Hemiptera: Corixidae) in the wetlands of Doñana National Park (SW Spain). Hydrobiologia 634: 209–217.CrossRefGoogle Scholar
- Sailer, R. I., 1948. The genus Trichocorixa (Corixidae, Hemiptera). The University of Kansas Science Bulletin 32: 289–407.Google Scholar
- Talling, J. F., & D. Driver, 1963. Some problems in the estimation of chlorophyll a in phytoplankton. In Doty, M., (ed), Proceedings of the Conference on Primary Productivity Measurement, Marine and Freshwater, pp. 142–146. US Atomic Energy Engineering Commission, Honolulu, Hi.Google Scholar
- Walton, M. E. M., C. Vilas, C. Coccia, A. J. Green, J. P. Cañavate, A. Prieto, S. A. van Bergeijk, J. M. Medialdea, H. Kennedy, J. King & L. Le Vay, 2015. The effect of water management on extensive aquaculture food webs in the reconstructed wetlands of the Doñana Natural Park, Southern Spain. Aquaculture 448: 451–463.CrossRefGoogle Scholar