Seasonal dynamics of larvae and adults of two Enochrus Thomson ( Coleoptera : Hydrophilidae ) species in temporary and permanent water bodies of an urban park in Buenos Aires

The dynamics of Enochrus variegatus (Steinheil) and Enochrus vulgaris (Steinheil) in temporary pools and permanent ponds were studied in an urban park of Buenos Aires, Argentina. Samples were collected from both types of habitats weekly throughout one year, and variables like water surface and vegetation cover were recorded. In permanent ponds, adults of both species were found throughout the study period, while in temporary pools their presence depended on the availability of water. Differences in their reproductive seasons and breeding sites were observed. Enochrus vulgaris was characterized by an opportunistic strategy, and its larvae were detected in both types of habitats, associated mainly with spring and summer temperatures. Conversely, E. variegatus larvae were recorded almost exclusively in temporary pools, during the winter and spring seasons. In permanent ponds, both species were associated with vegetation cover, which could be related to the availability of refuges from fi sh predators. First instar larvae showed the highest abundance and were captured more frequently, in consecutive weeks and simultaneously with third instar larvae, suggesting that the two populations studied are not synchronized in oviposition time or development of immature stages. In several occasions, second instar larvae of both species were captured in a pool that had been dry during the previous sampling date, suggesting that they had not hatched from fresh laid eggs. Our fi ndings show that adults of these species are capable of dispersal to and from temporary pools to avoid drought. In addition, their larval stages may also be able to resist drought in the pools by burrowing into the substrate.


INTRODUCTION
Lentic freshwater habitats in temperate regions range from small and ephemeral pools to large and permanent ponds or lakes that could be organized along a gradient of water permanence.Organisms inhabiting these habitats dif fer in their ecological strategies according to the range of the water permanence gradient they occupy.The main factors that determine the presence and success of a species within a specifi c portion of the duration gradient are physicochemical constraints and biological interactions (Wellborn et al. 1996).The permanence of water affects the relative importance of biotic and abiotic processes in determining the presence and distribution of species in freshwater habitats (Schneider & Frost 1996).
In habitats that frequently dry out, the main physical constraint for aquatic organisms is the lack of water.However, some organisms take advantage by colonizing these habitats, exploiting both the rich resources of fered by these temporar y pools and the relative absence of predators (e.g., fi shes).The aquatic invertebrates that colonize temporar y waters have developed different strategies to survive the dr y periods, which include physiological tolerance, effective capacity for migration, and/ or changes in life history (Wiggins et al. 1980, Williams 2006).Although the adaptations of different species are diverse, members of the same taxonomic group show similar strategies at the same stage of life cycle (Stanley et al. 1994, Williams 2006).
B o t h p h y s i o l o g i c a l a n d b e h a v i o r a l mechanisms to survive dry periods have been documented in aquatic coleopterans.In some cases, adults disperse oppor tunistically or seasonally to temporary bodies of water, where oviposition and development of immature stages occur.When the pools dry out, adults return to more permanent sites, where they find refuge over the unfavorable periods (Fernando 1958, Fernando & Galbraith 1973, Wiggins et al. 1980, Wellborn et al. 1996).In other cases, instead of dispersing to sites that retain water, adults may burrow into the substrate, either super ficially under dead leaves, or deeper in the ground (Williams 2006).
Although ecological studies about Enochrus are scarce, species of this genus are usually mentioned as par t of the aquatic insect community.Several studies have repor ted the presence of the genus Enochrus in water bodies of Buenos Aires (Fischer et al. 2000, von Ellenrieder & Fernández 2000, Fontanarrosa et al. 2004, 2009, Fernández & López Ruf 2006).Regarding immature stages, most of the information is published in the context of taxonomic descriptions of the species (Archangelsky 1997).No previous studies have been performed on the seasonal dynamics and breeding sites of lar val instars of Enochrus species, with the exception of a mention for E. vulgaris lar vae during summer in a permanent pond (Fernández 1992).The main diffi culty until now has been the impossibility to distinguish immature stages at the species level.Nevertheless, in a recent study we have described the larval morphology of E. vulgaris and E. variegatus (Byttebier & Torres 2009).This will allow the differentiation of immature stages of these species in field studies and thus represents an opportunity to investigate reproductive and ecological aspects of these species at the population level.
We aim to describe the seasonal dynamics of E. vulgaris and E. variegatus and their reproductive strategies in temporar y and permanent urban water bodies of the city of Buenos Aires.

Study area
The city of Buenos Aires is located in a temperate climate zone, with four clearly distinguishable seasons, mainly related to temperature differences.In this paper we follow the criterion of the National Meteorological Service, which classifies seasons according to the following three-month periods: December-February (summer), March-May (fall), June-August (winter), and September-November (spring).Mean temperatures are 23.6 ºC during the summer, 17.8 ºC during the fall, 11.5 ºC during the winter, and 17.3 ºC in the spring.Rainfalls occur year round, and although highest values are recorded in the summer (341 mm) and lowest values in the winter (199 mm), there is no distinguishable rainy season (National Meteorological Service 2011).
Field studies were carried out in the "Golf Club Lagos de Palermo", which covers 15 hectares, located in the north-eastern zone of the city of Buenos Aires (34º33' S, 58º26' W).The vegetation of this park is mainly grass subject to periodical cut, and isolated groups of trees.This recreational area contains fi ve permanent, human-made ponds with surface areas of at least 500 m 2 , and because of the irregular relief of the land, the formation of around 50 temporary pools with variable size and permanence is favored after rainfalls.Based on preliminary studies, we selected two permanent ponds and two temporary pools as representative habitats for the species analyzed in this study.
The permanent ponds were characterized by a constant water surface of 3000 m 2 (pond A) and 650 m 2 (pond B).Both ponds sheltered fi sh populations (Cnesterodon decemmaculatus [Jenyns]), which are potential predators of aquatic insects.Pond B lacked vegetation, while pond A was covered by fl oating (Azolla fi liculoides Lam.) and submerged fl ora.
Both temporary pools (pool C and pool D) showed fl uctuations in the water level, and dried out repeatedly through the study period.The maximum surface area and depth in each pool were less than 90 m 2 and 25 cm respectively.The substrate of both pools was covered by grass (similar to all temporary pools present in the studied area), and both pools were exposed to sunlight during most of the day.

Sampling and identifi cation
The permanent ponds and temporary pools selected were evaluated weekly throughout a one-year period (June 2001-May 2002).The maximum length and width (m) of each temporary pool were measured on each sampling date, and the proportion covered with water of the rectangle formed by these two variables was assessed.
Samples of aquatic insects were collected with a hand net (mesh size 300 µm, mouth opening 10.5 x 4.5 cm).The sampling effort in permanent ponds was constant (10 net sweeps covering 1 m each), whereas that in temporary pools was approximately proportional to the area covered by water on each sampling date, according to a previously adjusted scale (Fontanarrosa et al. 2004).The samples were fi xed in situ in ethanol 96 % and transferred to the laboratory for identifi cation.
In the laboratory, Enochrus spp.adult and larval specimens were identifi ed to species level by using the appropriate systematic keys and original descriptions (Fernández 1989, 1994, 1997, Oliva et al. 2002, Byttebier & Torres 2009).Adult specimens were identifi ed with a stereoscopic microscope, whereas larvae were identifi ed with an optic microscope.For each sample, adults and larvae of each species (E.variegatus and E. vulgaris) were counted.

Data analysis
Water dynamics and seasonal distribution of adults and larvae in temporary and permanent water bodies: For each sampling date and each temporary pool, the surface area was calculated by multiplying maximum dimensions by coverage proportion.In order to analyze fl ooding fl uctuations, water surface data of both pools were added by sampling date and transformed to percentage values, the date with the maximum fl ooding was considered as 100 %.Weekly values of percent surface area were compared between seasons with a Kruskal-Wallis test.Post-hoc multiple comparisons of mean ranks for all pairs of seasons were used to identify signifi cant differences (Siegel & Castellan 1988).Permanence of temporary pools was estimated as the sum of consecutive periods between sampling dates in which the pools contained water.After dry periods, the date of fl ooding of a pool was assigned to the date of the previous rainfall event.Meteorological data for the study period were provided by the National Meteorological Service.
Sampling dates were classified according to the season, and for each season the proportion of dates with adults and larvae of each species was calculated.On the other hand, the yearly number of sampling dates with adults and larvae of each species was divided by the total number of sampling dates to assess the overall proportion of detection in permanent and temporary waters separately.In the latter case, only dates when water was present were considered.The proportions obtained were compared between seasons (habitat types pooled) and between permanence categories (seasons pooled) by means of a chi square test for independent proportions (adults and larvae of each species separately).This test is comparable to computing the Pearson chi square statistic for contingency tables.Differences between pairs of seasons were examined by subdividing the contingency tables and computing the chi square value on the partial tables (Fleiss et al. 2003).Seasonal average of percentage fl ooding and number of collected adults and larvae of each species in ponds and pools were calculated for graphic display.For permanent ponds, which differed in vegetation cover, the proportion of dates when adults and larvae of each species were present was calculated and compared by means of the chi square test for independent proportions (Fleiss et al. 2003).
The sampling dates were also classifi ed into three categories of temperature ranges: < 15 °C, 15-20 ºC, and > 20 °C, according to the average temperature of the seven previous days.These temperature ranges were defi ned to be representative of winter, spring-fall, and summer seasons respectively.The number of dates with presence of adults and larvae of each species was divided by the total number of dates for each temperature range category to calculate proportions.These proportions were compared by means of the chi square test for independent proportions, and differences between pairs of ranges were examined by subdividing the contingency tables and computing the chi square value on the partial tables (Fleiss et al. 2003).Average abundances of adults and larvae of each species by temperature range were used for graphic analysis.The mortality rate of each instar was calculated for each species as the number of individuals that completed the instar divided by the number of individuals that reached the stage.

Larval dynamics in temporary pools
The presence and abundance of different larval instars of each species was analyzed in both pools.The time elapsed since flooding and the development time observed for each species in the laboratory were taken into account to estimate the expected larval instar for each sampling date.Dates when expected larval instars did not correspond to those observed in the fi eld were analyzed and described in detail.For these situations, average temperatures of the periods since flooding were calculated, and compared to those recorded in the laboratory during the development time estimations.

Water dynamics and seasonal distribution of adults and larvae in temporary and permanent water bodies
The water level in the permanent ponds remained constant.In contrast, the surface area of the temporar y pools showed pronounced fl uctuations over the study period.Signifi cant dif ferences of water sur face were detected between seasons (H 3 = 23.08;N = 51; P < 0.001).The lowest values were observed during summer (Fig. 1) and were signifi cantly lower than those observed during winter (P < 0.001) and spring (P < 0.001), when maximum fl ooding occurred.In fall, the water surface attained intermediate values and showed no signifi cant differences with the other periods.With regard to permanence, in general, flooding times were slightly longer in pool C than in pool D, although it varied seasonally in both cases.In pool C, the longest periods with water were 16 weeks in winter, nine weeks in spring, one week in summer and one to six weeks in fall (Fig. 2A).In pool D, water permanence was fi ve and six weeks in winter, fi ve and three weeks in spring, one week in summer and one to four weeks in fall (Fig. 2B).
Adults of E. vulgaris were collected in both types of habitats throughout most of the year, except during the driest period (summer), when they were recorded only in the permanent water bodies.Immature stages of this species were collected in both habitat types in spring and fall, only in permanent pools in summer, and no individuals were recorded in winter (Fig. 1A).Neither differences between seasons nor dif ferences between types of habitats were detected in the proportions of presence of adults or lar vae of this species (Table 1).Adults were detected in both permanent ponds, although mainly in pond A (which harbored both fl oating and submerged fl ora) (X 2 1 = 3.9; N = 102; P < 0.05), and larvae exclusively in this pond (X 2 1 = 6.38;N = 102; P < 0.05).With regard to temperature, adults of E. vulgaris were collected at the whole temperature range recorded during the study period in Buenos Aires and showed no defi ned pattern in relationship to temperature.Lar vae were collected at temperatures above 12.9 °C, and higher propor tions of dates and abundance were recorded at increasing temperatures (Fig. 3A).No statistical differences in proportions of dates with presence of adults or larvae were detected between temperature ranges.Relative abundances of adults and lar vae showed a pattern similar to the described proportions (Fig. 3A).
Adults of E. variegatus were collected both in temporar y and permanent habitats during most of the year, with the exception of the driest period, when they were recorded only in the permanent ponds (Fig. 1B).No statistical dif ferences in the propor tions of presence were detected between sites of dif ferent permanence, whereas signifi cant differences were detected between seasons (Table 1).The proportions of dates on which this species was present were higher in the winter-spring period than in the summer-fall period (X 2 1 = 9.08; P < 0.005).Immature stages were obser ved almost exclusively in temporary habitats (with the exception of one lar va during the spring season), and the dif ference of detection in

Species
Enochrus  Comparison of proportions of dates when the presence of Enochrus was recorded according to habitat permanence and season.
Comparación de la s proporciones de fechas con presencia de Enochrus de acuerdo a la permanencia del hábitat y la estación.
ns: non-signifi cant differences both types of habitat was signifi cant (Table 1).Larvae were collected only during the winter and spring seasons (Fig. 1B), when proportions showed signifi cantly higher values than during the summer-fall period (X 2 1 = 20.43;P < 0.001).
Adults were collected more frequently in Pond A, although statistical results were only marginally significant (X 2 1 = 3.46; N = 102; P = 0.06).The single lar va of this species collected in permanent water was also found in this pond.Adults were observed at the whole temperature range recorded during the study period (Fig. 3B).Differences in proportions of dates with presence were signifi cant among temperature ranges (X 2 2 = 9.38; N = 51; P < 0.01), with higher values at temperatures below 15 ºC (X 2 1 = 9.20; P < 0.005).Larvae of E. variegatus were captured only at temperatures below 20 ºC (Fig. 3B).The proportion of dates with detection of lar vae dif fered between temperature ranges (X 2 2 = 10.92;N = 51; P < 0.005), with signifi cantly higher proportions of dates with presence of larvae below 20º C than above this temperature (X 2 1 = 9.71; P < 0.005).Relative abundances of both adults and larvae showed the same trend as the described proportions (Fig. 3B).

Development time estimation
The development times obser ved for E. vulgaris were slightly slower than those for E. variegatus.The fastest instars for both species were the second lar val and the pupal stage (mean ± DE = 4.2 ± 1.48 and 3.27 ± 1.1).The third instar lar va was the slowest in average (10.83 ± 1.82) (Fig. 4).E. vulgaris hatched from the egg case on day 6 and reached the second and third lar val instar on days 13 and 16 respectively.Specimens moulted to pupa from day 26 on (Fig. 4A).Mortality for the fi rst larval instar  was 0.65, while no mor tality was recorded during the second and third larval instars, or the pupal stage.E. variegatus emerged from the egg case on day 3, reaching the second and third instar on days 9 and 13 respectively.The pupal stage started on day 21 (Fig. 4B).The mortality rates for the fi rst and third larval instar were 0.84 and 0.22 respectively, while no mortality was recorded during the second larval instar and the pupal stage.

Larval dynamics in temporary pools
The fi rst instar larvae of both species showed the highest abundance and were captured more frequently, while third instar lar vae recorded the minimum frequency and abundance.First instar larvae were observed in the fi rst week after fl ooding, and also several weeks after.They were also observed in consecutive weeks and simultaneously with third instar lar vae.The presence of the two fi rst larval stages of E. variegatus and E. vulgaris was observed on dates immediately before the pools dried out completely in several opportunities.
In two opportunities (October and April), the instar of E. vulgaris larvae observed in pool C did not match the expected one according to the fl ooding time and the development time recorded in the laborator y.On these dates, second instar lar vae were captured after the pool had dried out on the previous date.In October, the average temperature was 17.3 ºC and the time elapsed since the previous rainfall event was two days, whereas in April, these values were 18.0 ºC and six days respectively (Fig. 2A).
Similar obser vations wer e made for E. variegatus in pool D in three dif ferent oppor tunities.Second instar lar vae were collected on dates immediately after fl ooding in July, October and November.In July, the average temperature was 12.6 ºC and the time elapsed since the previous rainfall event was seven days, whereas in October the values were 17.3 ºC and two days respectively and in November 19.1 ºC and eight days respectively (Fig. 2B).

DISCUSSION
Lar vae and adults of both species were obser ved in temporar y pools and permanent ponds during the study period, with differences between seasons and type of water body in which they were recorded.
Results show that the r epr oductive seasons of E. variegatus and E. vulgaris dif fered through the study period, with a shor t overlapping period in spring.This differentiation is consistent with the association of each species to a par ticular temperature range.Enochrus variegatus reproduced at lower temperatures (especially during the winter season), while E. vulgaris was associated with higher temperatures (especially during the summer season).No previous data are available regarding the breeding season of these species of Enochrus to assess the inter-annual constancy of the pattern observed.
The presence of immature stages and adults of E. variegatus and E. vulgaris both in temporar y and permanent water bodies is consistent with other repor ts for the region (von Ellenrieder & Fer nández 2000, Fontanarrosa et al. 2004), and indicates good dispersal capacity of both species.This migration behavior is a common adaptation in insects that colonize temporary water bodies, and is also widespread among coleopterans (Wiggins et al. 1980, Williams 2006).
A m o n g p e r m a n e n t h a b i t a t s , t h o s e containing vegetation seem to represent more favorable environments for these insects, and the association of E. vulgaris and E. variegatus to aquatic vegetation has been reported previously in permanent ponds and marshes in central-eastern Argentina (Poi de Neiff 1983).The association of Enochrus with vegetation could be related to the availability of substrate for adults to attach their egg cases, as well as suitable places for larvae to rest and feed.On the other hand, vegetation may offer refuges to hide from predators, taking into account the presence of fi shes in both ponds studied.Experimental studies have shown lower colonization of Enochrus species in ponds that contained caged fi shes, as compared to fishless ones (Binckley & Resetarits 2005).Nevertheless, it is not possible to infer from this work whether adults avoid colonizing sites without vegetation, or whether they colonize them and low abundances result from predation pressure.Our results dif fer from those of another study in Buenos Aires, which repor ted higher abundances of lar vae and adults of Enochrus in ponds without vegetation as compared to vegetated ones (Fontanarrosa et al. 2004), but the presence of fi shes was not mentioned in this work.
The predominant presence of lar vae in temporar y pools indicates that they are favorable habitats for both species despite the unpredictable permanence of water.Our results suggest that E. variegatus prefers temporar y waters for reproduction.In contrast, the presence of E. vulgaris larvae was only slightly higher in temporary pools than in ponds.This might indicate a more opportunistic strategy of this species regarding the selection of breeding sites, especially taking into account that the reproductive season of E. vulgaris was coincident with the drought period during the study period.Dif ferent species of Enochrus have been reported to breed at different points of the permanence gradient, ranging from temporar y pools to permanent ponds with fi shes (Poi de Neiff & Neiff 1984, Fairchild et al. 2003, Fontanarrosa et al. 2004).Further research is needed to establish whether the strategy of exploiting temporary habitats differs between the species studied or whether it varies according to the environmental conditions (e.g., drought during the reproductive period).
The presence of first instar lar vae in temporar y pools over several consecutive weeks, and their coexistence with third instar lar vae indicate that populations are not synchronized in oviposition time and development of immature stages.These results and the detection of fi rst instar lar vae soon after the fi lling of the pools are consistent with an opportunistic strategy of both populations, exploiting temporary habitats for reproduction during the whole wet period (W illiams 2006).On the other hand, the relatively low abundances of the second and third lar val instars suggest high mor tality rates during immature development, with only a reduced propor tion of individuals reaching maturity.Nevertheless, it is not possible to assign the low survival in the pools to the effects of drought, since fi rst instar mortality of different Enochrus species has shown to be high in laborator y obser vations as well (this study, Hosseinie 1995).
The obser vation of second instar lar vae few days after the fi lling of the pools is an unexpected result, especially since laboratory data showed a longer development time (nine days for E. variegatus and 13 days for E. vulgaris to reach the second larval instar) even at temperatures higher than those observed in the fi eld.These results suggest the possibility that those larvae survived the drought and did not hatch from egg-cases laid after fl ooding, although we did not search for buried larvae to confirm this hypothesis.Never theless, a previous study of benthonic invertebrates from the fl oodplain of the Paraná River (Montalto 2008) reported the presence of lar vae of the genus Enochrus in soil samples during the drought phase of the wetland.
To our knowledge, this is the fi rst study in temperate Argentina, where the abundances of lar vae and adults are compared between temporar y and permanent aquatic habitats.The results suggest that the Enochrus species studied present two strategies to cope with habitat desiccation, which allow them to tolerate and/or avoid dry periods in the study region.
Future studies should assess the relative importance of tolerance/resistance (burrowing into the soil when the habitat dries out) and avoidance (migration to permanent habitats) of E. vulgaris and E. variegatus as strategies to deal with dry periods.
Data of development time and survival were obtained in laboratory conditions in the context of the taxonomic description of the larval stages (detailed description of methods in Byttebier & Torres 2009).Specimens of E. variegatus were raised from November 2006 to February 2007 (environmental temperature of 23-27 ºC), and E. vulgaris from April 2007 to June 2007 (environmental temperature of 19-22 ºC).After oviposition, egg cases and afterwards larvae were checked and fed, and the molts or deaths recorded daily.The average and range of duration of each larval instar was estimated, being the day of oviposition of egg cases considered as day zero.

Fig. 1 :
Fig. 1: Water availability (seasonal average fl ooded area [m 2 ] in temporary pools) and relative abundance (average number of individuals per sampling date) by season and type of habitat, of adults and larvae of Enochrus.The size of the circles is indicative of the relative abundance of individuals.(A) E. vulgaris; (B) E. variegatus.Disponibilidad de agua (área media inundada estacional [m 2 ] de charcos temporarios) y abundancia relativa (número medio de individuos por fecha de muestreo) de adultos y larvas de Enochrus por estación y tipo de ambiente.El tamaño de los círculos es representativo de la abundancia relativa de los individuos.(A) E. vulgaris; (B) E. variegatus.

Fig. 2 :
Fig. 2: Presence of three larval instars of E. vulgaris and E. variegatus in two temporary pools: (A) Pool C, (B) Pool D. Arrows indicate dates when second instar larvae were captured after the pool had dried out during the previous sampling.