FOOD COMPOSITION OF SOME LOW ALTITUDE LISSOTRITON MONTANDONI ( AMPHIBIA , CAUDATA ) POPULATIONS FROM NORTH-WESTERN ROMANIA

The diet of some populations of Lissotriton montandoni from north-western Romania is composed of prey belonging to 20 categories. The food components of the Carpathian newts are similar to those of other species of newts. Most of the prey are aquatic animals, but terrestrial prey also has a high percentage abundance. The consumed prey categories are common in the newts’ habitats as well, but in natural ponds the prey item with the highest abundance in the diet is not the most frequent one in the habitat. Thus, although the Carpathian newts are basically opportunistic predators, they still display a certain trophic selectivity.


INTRODUCTION
Lissotriton montandoni is endemic in the Tatra Mountains and in some areas of the Carpathian Mountains (Cogălniceanu 1997).A large part of its distribution area is in Romania, the Carpathian newts being present in the eastern Carpathians and in a very small area of the southern Carpathians (Cogălniceanu et al., 2000, Covaciu-Marcov et al 2009).In Romania it is present at altitudes higher than 500 m (Iftime 2005), but recently it has been recorded in the west of the country at 200 m (Covaciu-Marcov et al., 2007).The species is considered vulnerable, being affected by the deterioration of habitats (Iftime 2005).However, in Romania studies concerning the species' biology and ecology are missing.Lack of such knowledge is likely to affect any program for the protection of the species because understanding its habitat requirements and habitat use is a major challenge in conservationist biology (Hartel et al., 2006) and Europe is still in need of information about Romanian biodiversity (Hartel et al., 2008).
Feeding is an essential indicator of the position of a species in an ecosystem as it is the primary link between an animal and its environment (Kenett & Tory 1996).Quantitative information about the role of amphibians in ecosystems is extremely important in their capacity as consumers both in aquatic and terrestrial habitats (Whiles et al., 2006).In Romania there are no data about the feeding of Carpathian newts.A study on this subject was realized in Ukraine but was only on a few individuals and at an ordinary altitude for the species (Kuzmin 1990).Identification of the low altitude populations from north-western Romania raises the problem of the manner in which the species can find the necessary ecological conditions at those altitudes and if the trophic requirements are satisfied.
The purpose of our study lies in establishing the diet composition of some low altitude populations of L. montandoni from north-western Romania.We will address the following: 1) knowledge of food composition, 2) comparison between diet composition and prey availability, as was recently realized in a study on the benthonic fauna from the habitats of the newts (Cupşa et al., 2007), and 3) establishing trophic differences between the sexes, habitats and periods.

MATERIALS AND METHODS
The study was carried out during April-May 2005.One hundred and seventy nine Carpathian newts (101 males and 78 females), captured from two habitats, were analyzed..The newts were captured from an area located in the utmost north-western part of Romania, on the territory of Turţ Village, 10 km from the border with Ukraine.Near the locality were the mines from Turţ (Turţ Băi) but at present the area has only a few houses and abandoned mines.Thus, the current anthropogenic impact is reduced, but the effects of activities from the past are still evident.In the area there are buildings and abandoned equipment, heaps of barren rock and the brook is polluted even now due to the heaps of barren rock and decantation ponds.The newts' habitats are at 250 m altitudes and at about 500 m from each other, near an asphalted road which is little used.
The analyzed habitats will be referred to as Pond I and Pond II.Pond I is an artificial one that owes its existence to the mines.It is an excavation made close to a pipe from the mine.It has a length of 2 m, a width of 1 m and a depth of 1.5 m, the water level is relatively constant and the turbidity high.The vegetation is scruffy, consisting of a small amount of rushes on the shore.Near the habitat there is a coniferous plantation and then a beech forest.Pond II is natural and permanent, situated at the edge of a natural beech forest and at the base of a slope.It has a large surface, being a pooling area established at the level of a brook.The pond is 13 m long and 2 m wide, but it is surrounded by marshy areas which extend in rainy periods.The depth of the water is maximum 0.75 m, the bottom is covered by dead leaves and the shore by amphibious vegetation.
The newts were captured with the help of nets attached to long metallic handles or with a square dredge.The stomach contents were collected using the stomach flushing method (Solé et al., 2005), and the newts subsequently released.Samples were stored in test tubes, preserved with formaldehyde and examined in the laboratory.Food composition was evaluated by percentage abundance (%A) and frequency of occurrence (%f).Dietary diversity was estimated with the Shannon-Wiener (1949) diversity index (H), and dietary similarity with the Sorensen index (Chao et al., 2005).The Kruskal-Wallis and Mann Whitney U test was applied for comparison of the data sets (Kruskal & Wallis 1952).

RESULTS
Not all of the studied Carpathian newts had stomach contents (Table 1).The individuals that did not feed make up a relatively important part of the total populations and were especially numerous in the artificial habitat (Pond I).The highest number of newts without stomach content was caught at the beginning and at the end of the aquatic period.In the middle of the period such individuals were missing or were very rare.The number of prey/individuals exhibits great differences both between habitats and between periods and sexes.The lowest values, both for the maximum and average number of prey/individual were recorded in the artificial habitat (Pond I).The same was true of the occurrence of empty stomachs: the least prey was consumed at the beginning of the period of study.The number of prey rose towards the middle of the aquatic period and then decreased again (Table 1).
Five categories of stomach contents were identified in the case of the Carpathian newts that had fed: 1) prey of animal type, 2) remains of plant materials, 3) fragments of molted skin, 4) amphibian eggs and 5) inorganic particles.The two populations of L. montandoni had consumed a total number of 462 prey belonging to 20 prey categories.From the stomach contents of the newts captured in Pond I only 72 prey items were identified and the remaining 390 prey items were collected from the newts from Pond II.The diet of the population from Pond I consisted of 11 prey categories and that from Pond II comprised 16 prey categories.In both habitats the number of the consumed prey categories displayed the same evolution as the number of prey/individuals, and was lowest at the begin-ning of the aquatic period, increasing subsequently and decreasing again at the end.Of the 20 prey categories, 7 were consumed in both habitats (Table 2).In Pond I none of the prey categories was constantly used in each period.However, in the third habitat, 3 prey categories (Gastropoda, Ephemeroptera and Plecoptera larvae) occurred in all sampling periods, plecopteran larvae being present in the stomach contents of both sexes in every period.The percentage abundance and frequency of occurrence of prey exhibit variations usually both between periods and sexes.
Generally, the prey categories with high percentage abundance also had a significant frequency of occurrence.In Pond I the most important prey categories are different comparing the three periods, but in Pond II plecopteran larvae exhibit very high values both for percentage abundance and frequency of occurrence in all sampling events (Table 2 and 3).
The plant materials and molted skins of amphibians were constantly consumed in both habitats and by both sexes.Eggs were identified in the stomach contents of a few newts from both habitats but only in certain periods.Inorganic fragments occur accidentally and exclusively in Pond II and only in some of the periods.Dietary diversity is variable between sexes and habitats, but these variations follow the pattern indicated by the empty stomachs and the number of prey/individuals.Thus, food diversity is lower for Pond I than for Pond II, as well as at the beginning of the study (Table 1).
The results of the Mann Whitney U test show that the differences between the two sexes and the two habitats are not significant (P>0.05).For analyzing the differences between the periods we used the Kruskal Wallis test, its results showing insignificant variations between different months (H=5.67,df.=2, P>0.05).

DISCUSSION
L. montandoni feeds less intensely than other species of newts found in Romania (David et al., 2008a, b, Dobre et al., 2007, Ovlachi et al., 2007).The number of empty stomachs is very high and they are present during almost the entire aquatic period, and the number of prey/individuals is low.The feeding intensity of newts is lower at both the beginning and the end of the aquatic period (Covaciu-Marcov et al., 2003, Cicort-Lucaciu et al., 2006) due to the adaptation to this environment and then to the loss of it.Usually, amphibians had a low rate of feeding intensity at the beginning of their activity period (Hirai & Matsui 2000;Kovács et al., 2007) due to unfavorable weather conditionsmainly low temperatures which affect both the predators and especially the prey (Guidali et al., 1999;Covaciu-Marcov et al., 2003).This explains the feeding rate from the first period when indeed the temperatures were low.
Generally, it is considered that the situations in which the entire population had fed indicate optimal feeding conditions (Sas et al., 2009).Thus it seems that in the habitats from Turţ the Carpathian newts do not have favorable feeding conditions or are possibly less adapted to forage in the aquatic environment compared to other newt species.This might be a consequence of its adaptation to higher areas where difficult conditions lead to the shortening of the aquatic period and influence feeding in this environment.Nevertheless, the above-mentioned suppositions cannot be stated without reservations because at Turţ L. montandoni is recorded at lower altitudes than in other areas from Romania (Covaciu-Marcov et al., 2008, Gherghel et al., 2008).
In Pond I the feeding conditions are even more difficult due to the anthropogenic impact on the habitat.As this pond is artificial it does not have a well-defined zoocenosis, and thereby no potential preys.This fact is also supported by the benthonic fauna of the two habitats, there being only 2 taxa in Pond 1 compared to 9 in Pond II (Cupşa et al.,  2007).These data confirm that the trophic diversity depends on the quality of the habitat inhabited by amphibians (Kovács et al., 2007).The anthropogenic impact on the population of the artificial habitat becomes obvious, knowledge about feeding being crucial for understanding the effects of habitat modification on amphibians (Anderson et al., 1999).
The changes in time of the feeding intensity follow the evolution of climatic conditions.The increase of temperature increases feeding intensity and the newts consume more prey.A similar tendency was previously reported for other species of newts (Covaciu-Marcov et al., 2003), the differences being determined the variation of prey abundance between different periods (Fasola & Canova 1992, Denoël & Andreone 2003).However, the maximum number of prey/individuals (only 11 items) is significantly lower than for other species of newts (Cicort-Lucaciu et al., 2007, David et al., 2008b, 2009).This is a consequence of the consumption of relatively large sized prey.From the diet composition of the L. montandoni populations small prey (eg.Crustacea: Copepoda, Ostracoda or Cladocera) which are responsible for the high number of prey in the case of other species of newts, are absent.But this situation is an outcome of the conditions from the two habitats from which the microcrustaceans are missing (Cupşa et al., 2007).Probably their absence is a consequence of the colder microclimate which is obvious in the case of Pond II which is shaded and situated at the edge of a forest.
If there are microcrustaceans in the water, they are consumed in large quantities by the Carpathian newts (Kuzmin 1990) as well as by the other species of newts (Joly 1987, Schabetsberger et al., 1995, Kutrup et al., 2005, David et al., 2007).At Turţ, this is the case of a pond which was not taken into account because it had dried up by the end of April, where newts had massively consumed microcrustaceans which formed important agglomerations in some sectors of the pond.In that pond the maximum number of prey/newt was 96, much higher than in the other two habitats.These data prove that the Carpathian newts change their diet compo-sition according to the abundance of prey from the habitat.
The differences between the two sexes are not significant, but these differences affect the feeding intensity as well as the number, percentage abundance and frequency of occurrence of prey items.In Pond III the feeding intensity is always higher in females; they consume more prey categories (Table 2).This was predictable since in the breeding period the males, even hungry, are more interested in the nuptial parade than in feeding compared to the females (Ranta et al., 1987).In other amphibian species the feeding of females in the breeding period was also more intense and they consumed more large sized prey than the males (Juncá & Eterovick 2007).Nevertheless, in Pond I, apart from the last period, the situation is reversed, probably as a further consequence of the artificiality of the pond which alters the results due to the reduced prey availability.
Although apparent, not even the differences between habitats are statistically significant.It is clear that Pond I does not provide optimal trophic conditions for newts.Their feeding is deficient, which is evident from the lower percentage abundance of aquatic prey which reflects the lack of the proper trophic resources of the pond (Table 1).The trophic diversity is lower in this habitat.Thus, for the population of L. montandoni, breeding in the artificial habitat is probably a solution of the crisis caused by the lack of trophic resources.Nevertheless, the habitat is permanent and free from predatory invertebrates (Cupşa et al., 2007), a fact that probably favors the newts, which have been recorded in this habitat for many years, despite the fact that the natural pond is only at an altitude of 500 m and in other species of newts migrations up to 2000 m have been registered (Schmidt et al., 2006).
In the aquatic period, the newts mainly feed on aquatic prey (Kutrup et al., 2005, Dobre et al., 2007).Nonetheless, in our study the terrestrial prey have relatively high values, higher than those achieved by many other populations of newts.The differences between habitats are maintained concerning the origin of prey as well; in Pond I the terrestrial prey is more numerous.Thus, the trophic scarcity in Pond I that forces newts to forage in the terrestrial environment in order to complete dietary requirements is highlighted.This situation, where a habitat with low trophic offer forces the newts to forage massively in the terrestrial environment, has been observed in other species such as Triturus cristatus (Covaciu-Marcov et al., 2002a, 2003).
The consumption of terrestrial prey by newts in the aquatic period is considered to be a consequence of the fortuitous presence of these prey items in the aquatic habitat where the newts capture them from the water surface (Denoël et al., 1999, Denoël & Joly 2001).This is acceptable especially in the case of flying prey like flies.The consumption of other terrestrial prey (ants and pseudoscorpions) seems to indicate that some individuals of L. montandoni leave the water and forage in the terrestrial environment.This is plausible, considering that after the breeding period, the newts feed in the terrestrial environment.It is hard to believe that ants, or indeed pseudoscorpions which have low mobility and spend a lot of time in hiding, enter the water accidentally in large numbers.Pseudoscorpions represent a primary prey item in the diet of newts from Romania and are sometimes identified in the food composition of other amphibians (Groza et al., 2008).Thus, due to the low trophic offer, the newts, at least those from Pond I, sometimes forage in terrestrial environment.Feeding escapades into the terrestrial environment have been suspected for other newt species as well, e. g.Messotriton alpestris (Kuzmin 1990).
Plant materials in the stomach content is constantly encountered in amphibians (Santos et al., 2004), their consumption being considered accidentally (Gunzburger 1999, Kovács et al., 2007).Vegetal remains have a relatively high frequency of occurrence in both sexes in each period.Similar to plant materials, the remains of molted skin are frequently consumed by a significant proportion of the Carpathian newts.These items are regularly present in the stomach content of newts, but gene-rally with much lower frequencies (David et al., 2007, Ovlachi et al., 2007).The high frequency of the occurrence of molted skins, especially in Pond I, strengthens the idea of the existence of inadequate feeding conditions in this habitat.
Amphibian eggs have been consumed in both habitats, with higher frequency in Pond I at the first sampling event.At that time there was an agglomeration of egg clutches of Rana temporaria.These have provided a consistent trophic basis for an important number of newts, a fact which is understandable in a small habitat, and this kind of trophic element is probably identified on an olfactory basis (Joly 1981).Nonetheless, the consumption of eggs did not reach comparable values with those recorded in other species (Covaciu-Marcov et al., 2002b, 2003, Cicort-Lucaciu et al., 2006).Perhaps, at that time some newts foraged in the terrestrial environment and the others, due to the low temperature.were less active and stayed hidden under the vegetation near the shores.Eggs were consumed in the first period in Pond II, but by a smaller number of newts.Due to the larger area of the pond, the newts had more difficulty in finding the eggs of Rana temporaria, as these are laid on the shores in sectors with little water.In the last period, eggs (in this case newt eggs) were consumed only in Pond II.The intake of amphibian eggs is beneficial for several reasons one of which is the low energy consumption needed for a large volume of food (Denoël & Demars 2008).Ranid larvae were consumed in both habitats, but display high percentage abundance in Pond I in the middle of the aquatic period of newts.The larvae proceeded from Rana temporaria spawns.Being abundant in a small habitat, the larvae comprised a quantitatively important resource which is very useful in this artificial habitat.
Despite the reduced feeding compared to other newt species, the Carpathian newts from Turţ consumed twice as many prey categories as those from Ukraine (Kuzmin 1990).This has several explanations.Firstly, the 10 newts from Ukraine were too few to provide an adequate estimation.Secondly, the samples were taken only once, toward the end of the aquatic period (Kuzmin 1990).The additional prey categories from Turţ are represented by aquatic mollusks (gastropods and lamellibranchiates), but most important by terrestrial prey (spiders and ants) which customize the habitat with reduced prey availability.Also, there are great differences between the percentage abundance and the frequency of occurrence of prey compared to the previous study (Kuzmin 1990).In Ukraine, most of the prey were microcrustaceans which were missing in the case of the populations studied by us.
Comparing the prey availability from habitats (Cupşa et al., 2007) with the diet of newts it can be observed that generally the prey consumed intensively by newts are important elements of the pools.The situation is even more obvious in Pond I, where the dominant categories from the habitat are prevailing in the diet too.This is once again a consequence of limited prey availability; in this habitat the newts do not have the possibility to choose between prey.In Pond II the situation is slightly more subtle.In the diet of the newts from here plecopteran larvae are predominant, but although they are important elements of the habitat they never attain highest abundance.Regarding the prey availability of the habitat, the most numerous are the gastropods (Cupşa et al., 2007).
It is possible that at least in relatively large and ecologically complex natural habitats, the trophic opportunism of newts is more gradated.The size of a pond also conditions the ability of the prey to avoid predators.Gastropods are present in the diet of newts as well, but with low percentage abundance although their frequency of occurrence is sometimes quite significant.Probably gastropods, being slow-moving animals, are less accessible.As they gather at the substratum level, adhering to different structures and they are located on sides and are inaccessible to newts.It is possible that snails are difficult to detach from the aquatic supports to which they adhere.Therefore it is not reduced mobility but the possibility of being ingested that matters.On the contrary, the plecopteran larvae have larger sizes but no shells and are therefore easier to digest.It is considered that although newts are opportunistic predators (Covaciu-Marcov et al., 2003, Ferenţi et al., 2008) they still exhibit a certain level of trophic selectivity as revealed by previously recorded cases where newts did not always consume the most abundant prey in the environment (Covaciu-Marcov et al., 2002b).There is a form of selection according to accessibility, size and utility of prey (Joly & Giacoma 1992) but also a selection made of the most numerous prey (Cicort-Lucaciu et al., 2005, 2007).Our results support this point of view, and though L. montandoni is an opportunistic predator, it still performs a certain selection between the most accessible prey from a habitat, hence a kind of controlled opportunism.

Table 1 .
The percentage of empty stomachs, feeding intensity, origin of the preys, food diversity (Shannon Wiener index) and similarity (Sorensen index), and frequency of vegetal fragments, shed skin,eggs and minerals (M-males, F-females, IV-April, V-May)