Benefit of semiaquatic mustelids from beaver construction activity in Belarus and the method to census aquatic prey

Having pronounced construction instinct and activity, beavers change a lot their aquatic habitats in order to provide foraging pathways protected by water environment and create effective shelters saving them from enemies (mainly wolves and humans) and cold weather during overwintering, giving birth and raising a litter as well as  everyday resting.
With respect to semiaquatic mustelids i.e. otter and mink, a question arises do these changes in aquatic habitats bring benefit for them or not?  Beavers build a lot of shelters in kinds of burrows and lodges, and such a benefit of better sheltering environment for otter and mink is evident. While investigating the question, we were mainly interested in beaver activity-related increase of water-dwelling prey of otter and mink in aquatic ecosystems and, first of all, in small streams such as small rivers, brooks and drainage canals. Just at small watercourses such an effect of damming by beavers on semiaquatic mustelids may be the most pronounced. By building dams, beavers create ponds; such ponds are gradually eutrophicated and densely overgrown with macrophytes. So, it was essential to get know, such beaver ponds bring benefit in aquatic prey supply for semiaquatic mustelids or not, and if it is,  on which stage of the pond eutrophication such benefit is the highest.

This slideshow requires JavaScript.

In the late 1990s and early 2000s me with my research group on vertebrate predators  in Institute of Zoology NAS of Belarus studied on the questions in Belarus (e.g. Sidorovich, 2011), and I evaluate that study was  one of our research lucks. I would like to present the study once more in this post (see for the details in my vertebrate predator monograph by Sidorovich (2011)).

First, I need to present our materials characterizing beaver inundation at small watercourses, where possible positive effect of beaver ponds on semiaquatic mustelids may happen. Area of 194 measured beaver ponds in Belarus in 2000-2003 varied up to 164100 m2, and on average constituted about 5101 m2: 70 ones in small rivers – 260-164100(5246) m2, 41 ones in brooks – 120-124800(13560) m2, and 83 ones in drainage canals – 20-13880 (800) m2. The largest beaver ponds were found at brooks. The mean area of beaver pond per 1 km of small stream stretch was as follows: small rivers – up to 13446, an average for this habitat type was 4094 m2/km, brooks – up to 120635 (21056) m2/km, drainage canals– up to 2756(558) m2/km. Inundation by beavers of brooks were the largest compared to that of small rivers (about 5 fold larger) and drainage canals (about 38 fold larger), especially if to take into account only brooks populated by beavers, where the average inundation of brook habitats constituted 31583 m2/km.

This slideshow requires JavaScript.

As to succession stage of beaver ponds relating eutrophication with an age, 6% of them were recent inundations, 26% of ones were somewhat eutrophicated and overgrown by macrophytes, 43% – moderate stage, 18% – heavy overgrown by macrophytes and much eutrophication, and 7% – stage of grassy marsh with some open waters.

In effect of damming in Belarus, beavers inundated up to 60.3, on average 10.6% of small watercourse valleys. Shallow waters (up to one meter) prevailed in beaver ponds. In beaver ponds located in small rivers the ration littoral parts and relatively deep waters comprised from 0.4 to 4.2, on average was about 1.7 fold, in brooks – 0.9-15.0(6.2) fold, in drainage canals – 0.3-2.1(1.0) fold. The proportion of relatively deep waters (deeper than  one meter) in the beaver ponds investigated varied from 4.4 to 61.6%, and the mean was 33.7%.

By investigating influence of beaver inundation on habitat carrying capacity (first of all, on aquatic prey supply) and population density of otter and mink on small watercourses, the following questions were raised. The main questions were as follows:

–  what are the changes in fish species diversity, population density and biomass in connection with inundation by beavers and eutrophication of beaver ponds with age;

– does creation of ponds by beavers lead to an increase in density and biomass of other prey categories of semiaquatic predators such as crayfish, rather big aquatic beetles, edible frogs, and how their abundance is changed with age eutrophication of beaver ponds;

– how does hydrochemical specificity of beaver ponds affect their productivity in relation to fish and other prey;

– what is the increase in aquatic prey after creation of pond by beavers, which can be calculated by comparing whole beaver pond and the neighboring non-flooded stream section of the length equaled to one of the stream part inundated by the pond;

– what is the difference in density or unit biomass of aquatic prey between the pond and stream calculated for 100 m2 plots;

– what is the changes in semiaquatic mustelids  in connection with inundation by beavers and eutrophication of beaver ponds with age.

Here and in the mentioned book (Sidorovich, 2001) I tried to answer the questions raised so far as it was studied.

To estimate the species diversity, density and biomass of aquatic prey of otter and mink in beaver ponds in comparison with non-inundated parts of the same small watercourses, we (me and my research group) created a special net equipment; several nets, which were adapted for size, flowing rate and other parameters of watercourses, were constructed, but mainly two kinds of net equipment were used.

This slideshow requires JavaScript.

The net equipment used for slowly flowing (slower than 0.3 m/s) small rivers of 7-25 m wide and 1-2.5 m deep was named ‘big net equipment’. It consisted of two net walls (15-25 m wide, 3-4.5 m deep, approximately). The net walls were fixed by wooden stakes and ropes as well as provided with a heavy chain fixed by rope at the lower side. One of the net walls had large catching cavity (hereafter ‘catching net’). The catching cavity looked like a 7-10 m fyke with a funnel-shaped entrance. Its mouth was located about 50 cm above the lower net side and had an oval shape of 4-7 m at its widest point and 1-1.5 m at the narrowest point. While using big net equipment to do census of water-dwelling prey in a watercourse section, the catching net was situated across the watercourse in the downstream part of the section. Auxiliary relatively small nets were used to shut all holes through which aquatic prey may escape out of the section. A kayak was necessary to fix the net on the opposite side of watercourse, if it was too deep to walk across. Pulling with a rope was useful to get the net equipment onto the opposite side of watercourse. The other net wall of big net equipment (hereafter ‘chasing net’) was also situated across the watercourse, in the upstream from the location of the catching net. Then the chasing net was gradually moved by at least two, preferably three or four people, from the upstream to the downstream part of the section. People pulled the net using wooden stakes and ropes, scraping along the banks. Other people in wet suits cleared wood material from the bottom in the section. This action scared aquatic prey, directing many of them to the fyke in catching net. It was also important to catch aquatic prey with landing nets in the section. This made more prey to move towards the fyke, and part of prey was captured by landing nets. This was especially important when two net walls were close to each other. The total time necessary to accomplish this kind of census of water-dwelling prey in a 150-300 m section of slowly flowing and relatively deep and wide small watercourse varied between 3 and 7 hours depending on the number of people involved (4-7 men), the length of stretch chosen and the complexity of the watercourse structure, i.e. bank type and abundance of dead tree material and aquatic plants.

This slideshow requires JavaScript.

Census of water-dwelling prey in small watercourses with higher flowing rate (faster than 0.3 m/s) was similar, but details varied in the procedures and equipment. Such small watercourse was usually 5-16 m wide and 0.5-2 m deep. The net equipment  was smaller (hereafter ‘small net equipment’), and consisted of two similar constructed net walls 8-13 m wide and 1.5-2.5 m deep. One of the net walls has a relatively big fyke with a length of 2-5 m and an oval mouth size of 1.5-2 m by 0.6-1 m (referred to as ‘moving catching net’). The fyke mouth was located at the lower net side. When using small net equipment to do a census of water-dwelling prey in a watercourse section, the moving catching net was situated across the watercourse in the downstream part of the section. The other net wall of the small net equipment had no fyke when the stream was too small, or had one or two fykes (hereafter ‘fixed catching net’). It was fixed across the watercourse in the upstream part of the location of the moving catching net. All small holes through which aquatic prey may escape from the section were shut. Then two, preferably three, people gradually move the moving catching net from downstream to upstream in the section. The total labour time required to complete census of water-dwelling prey in a 100-150 m section of this watercourse varied between 3 and 6 hours depending on the number of people involved (3-5 men), the length of the stretch chosen and the complexity of stream structure. Usually, at least 3 hours were spent for census of aquatic prey in a 100 m stretch of a small watercourse flowing faster than 0.3 m/s.

 All nets were made from cord of 2 mm in diameter. The mesh of both net types was 6 mm. Only very small aquatic animals, that were usually not important as possible prey, could get through the net. Rate of underestimation of recorded prey was assessed by repeating the census method on the same watercourse stretch: two extra guarding net walls were fixed across the river at the section borders and the catching was repeated several times. Our experiences showed that two repetitions were sufficient, because either no aquatic prey or only a few very small ones were captured in the second repetition.

This slideshow requires JavaScript.

To estimate the species diversity, density and biomass of water-dwelling prey in beaver ponds, and the increases in these variables while compared to watercourse part unaffected by beaver damming, the following method again based on net equipment was applied. First, the area of a given beaver pond in its shallow and relatively deep waters should be assessed. Formally, depth of one meter was established as the border between the shallow and deeper waters. In practice, the depth in different parts of beaver pond was measured by walking and swimming or by special measure-stick in ice-bound period. Also, to differentiate shallow and deep waters in beaver pond, the distribution of aquatic plants normally growing densely in littoral parts was taken into account. After mapping, it was not hard to measure the area of the beaver pond and areas of its shallow and deep waters. Second, a section census of water dwelling prey in the beaver pond was carried out. Usually, up to 8 sections of 15 m2 each were investigated for littoral part and deep waters. Sometimes, it was possible to apply in beaver pond the same net equipment as the one described above used in watercourses. Respectively, census section sometimes was much bigger – up to 500 m2. Special test that was fulfilled for several beaver ponds suggests that pooled area of census section should cover approximately 1% of large beaver ponds (> 10000 m2), 2-10% of medium-sized ones (1000-10000 m2), and 30-50% of small ones (<1000 m2).

The net equipment applied for census of water-dwelling prey in sections of 15 m2 consisted of two continuously standing net walls (5 m wide, 3 m deep) which were located across the beaver pond at a distance of 3 m. Additionally, other two moving net walls (3 m wide, 3 m deep) were used. All these net walls were fixed by wooden stakes and ropes. By fixing the net walls around a particular section of beaver pond, the section bottom was not destroyed. After the net equipment was fixed it was left open for several hours and then rapidly shut. Then the water-dwelling prey contained in the net enclosure were captured by use of landing nets until all were caught. This was defined by a zero asymptote in the catch result for each prey category (fish, crayfish, frogs and aquatic beetles). The landing nets were 0.7 m in diameter and have a mouth of 0.3 m in diameter. The species composition of water-dwelling prey in beaver pond was assumed to be their total diversity in all the sections studied. The average biomass of a given prey species or category were calculated for shallow and deep waters separately, and then these mean values were extrapolated for the whole beaver pond. The third important thing was estimation of length of the stream part flooded by the beaver pond. It was not hard to reveal the former location of stream looking at tree distribution within beaver pond because bigger trees usually grow at banks. So, after the third step was completed and having investigated the adjoining watercourse part that was not inundated by beavers, there are all necessary parameters to compare the total number and biomass of aquatic prey in beaver pond and in the respective stretch of watercourse before damming by beavers (equaled the stream part inundated by the beaver pond). By comparing a unit plot (e.g. 100 m2) of beaver pond with the same water area of non-inundated adjacent part of the small watercourse, relative values of the change in aquatic prey may be assessed.

Now about the results that were gained through the study.

In the cold season, the aquatic prey biomass in beaver ponds varied up to 13 and on average was 1.6 kg per 100 m2 (n=20 beaver ponds investigated). That included fish, crayfish, rather big aquatic beetles, and hibernating frogs (mostly the common frog) which were captured in beaver pond under the water area of 100 m2. In the warm season, the aquatic prey biomass in beaver ponds (n=23) was found to be markedly lower than that in the cold season – up to 3.9, mean 0.6 kg per 100 m2. This essential difference between the seasons (on average 2.7 fold) mainly appeared due to wintering concentrations of common frogs in part of beaver ponds investigated. In such beaver ponds common frogs constituted 26-91(74)% of the prey individuals censused and 29-97(86)% of the prey biomass. It was from 0.2 to 11.3, on average 3.8 kg of common frogs per 100 m2 of such beaver ponds. Common frogs tended to hibernate in rather young (0.5-6 years old) and quite small beaver ponds, and nearly in a half of the beaver ponds investigated (11 out of 20, 55%) only few common frogs were found – less than 10 per 100 m2. Usually such beaver ponds were characterized by very low oxygen concentration – lower than 4 mg/l.

Concerning the change of aquatic prey biomass in beaver ponds in relation to their eutrophication with an age, the following data were obtained. In the warm season aquatic prey biomass in beaver ponds, first, increased in succession gradient and then decreased, because the same trend attributed to fish as main aquatic prey component. Again the low oxygen concentration and too high abundance of sulphuretted hydrogen was plausible explanation for the decline. In the cold season the biomass of water-dwelling prey was negatively correlated with the eutrophication of beaver ponds investigated. This correlation was based on the revealed sharp decreasing of common frog numbers with the increasing of beaver pond eutrophication.

By comparing a unit plot (e.g.100 m2) of beaver pond with the same water area of non-inundated adjacent part of the small watercourse, it was found that the aquatic prey biomass was mainly higher in beaver ponds: in the cold season – 3 fold on average, in the warm season – 15 fold on average. Nevertheless, parts of the studied beaver ponds were less productive than the adjacent non-flooded stream section: in the cold season – 40.0% of the ponds investigated (n = 20 ponds investigated); in the warm season – 13.0% (n = 23). The total biomass of aquatic prey in beaver pond, however, was much higher than that in the stream stretch (i.e. the stream part inundated by the beaver pond) before damming by beavers, there. We found this by comparing the total biomass of aquatic prey in beaver pond with that in the adjoining non-inundated stream stretch of the length equaled to one of the stream part flooded by the beaver pond. In the cold season, it was up to 1802 (mean 127) fold higher, in the warm season – up to 1403 (mean 372) fold higher. This suggests that such an increase in aquatic prey biomass took place in stream part inundated by beavers since pond was created there.

Concerning crayfish, the following trends were revealed. First, the crayfish species evidently avoided eutrophicated beaver ponds having sulphuretted hydrogen deposits. Secondly, if the crayfish species is present in a small watercourse, it inhabited densely slightly eutrophicated beaver ponds located at the small watercourse, and the unit biomass was higher than that in the neighbouring part of the stream, but not inundated by beavers. Among such a little eutrophicated beaver ponds there were only recently created ones and rather small ones with well running waters.

Concerning fish species richness and biomass, the following data were obtained. In non-flooded parts of 18 small watercourses investigated (6 small rives, 7 brooks, and 5 drainage canals), a total of 21 fish species were recorded: small rivers – up to 18, mean 12.2; brooks – up to 11, mean 6.9; drainage canals – up to 10, mean 4.1 fish species. Pronounced difference in fish species diversity dwelling in small rivers along the flowing rate gradient was revealed. In fast flowing (>0.6m/sec) small rivers up to 12 fish species were recorded in census captures, in moderately flowing ones – up to 16, and in slowly flowing (<0.3 m/sec) ones – up to 18. In non-eutrophicated beaver ponds that were recently created or with flowing waters fish species were almost the same as those in the adjacent non-inundated watercourse parts. Fewer or the same number of fish species were found in slightly eutrophicated beaver ponds. Marked changes in fish community were registered for heavy eutrophicated beaver ponds, which consisted of disappearance of fish species characterized by strict requirement to a high oxygen concentration. Those were the troat, burbot, riffle minnow, minnow, and also bleak, dace, gudgeon and stone loach. Together with this decline in the fish species richness, several fish species attributable for non-flowing eutrophicated waters appeared in such beaver ponds. The species were the verkhovka, crucian carps, tench, and mud loach. Part of fish species were remained in the fish community despite of eutrophication. Those were the pike, perch, roach, three-spined stickleback and spiny loach. In heavily eutrophicated beaver ponds largely overgrown by macrophytes fish species diversity decreased, and maximally 6-9 fish species were registered. Among these species there were two different groups. First one consisted of species that well adapted to live in the conditions of low oxygen level. Those were the verkhovka, crucian carps, tench, and mud loach. The other group mainly includes wide-spread predator species such as the pike and perch. Sometimes in eutrophicated beaver ponds getting much waters from underground sources and having quite high oxygen concentration the fish species composition may be rather diverse including species groups which are attributable for either fast running waters (the troat, burbot, riffle minnow, minnow, bleak, dace, gudgeon, and stone loach) or eutrophicated non-flowing aquatic ecosystems (the roach, ide, common bream, cracian carps, tench, rudd, verkhovka, and mud loach) or widely spread fish species such as the pike or spiny loach.

In the cold season, the biomass of fish in beaver ponds varied up to 2.9, on average was 0.4 kg per 100 m2. At first, it tended to increase with eutrophicating, but then in heavily eutrophicated beaver ponds the fish biomass declined. Similar values of fish biomass (maximum – 3.8, mean – 0.5 kg per 100 m2) and pattern in relation to eutrophication were recorded in the warm season. The total biomass of fish in beaver ponds was much higher than that in the neighboring stream part of comparable length. In the cold season, it was up to 287 (mean 62) fold higher; in the warm season – up to 223 (average 76) fold higher. This suggests that such an increase in fish biomass took place in stream part inundated by beavers since pond was created there.

Thus, appearance of beaver ponds leads to increasing in aquatic prey biomass mainly due to growing of fish biomass. In the cold season, this trend is also supplemented with hibernating concentration of common frogs, but only in slightly eutrophicated beaver inundations or rather small ones having well running waters.

An important thing was to estimate increase of aquatic prey biomass in the whole stream valley due to damming by beavers. In the valleys of 25 model small watercourses up to 60%, on average 11% of them were inundated by beavers, and the water area increased up to 135, mean 9 fold. Rough estimates of aquatic prey biomass increase after the beaver damming were as follows: in the cold season – up to 984 (mean 32) fold, in the warm season – up to 3117 (mean 131) fold compared to a hypothetical situation without beavers.

It is noteworthy to pay special attention to edible frogs in beaver settlements located at small watercourses. Favourable conditions for spawning, sufficient exposure and good food supply attract numerous edible frogs to beaver ponds. In the warm season, their distribution in beaver ponds was evidently determined by difference in exposure. Usually edible frogs stayed in less shaded and well-warmed parts of beaver ponds. Average density of edible frogs in these habitats was strongly related to the age of beaver ponds. Concerning succession of beaver ponds due to their eutrophication with age, the following trends in edible frogs were recorded. At first, edible frog density mainly increased with the age of beaver ponds. This trend was going for approximately 15 years. It was plausibly connected with better exposure in aged beaver ponds due to dying of the flooded trees and bushes. Then edible frog density gradually declined. This opposite trend is also may be interpreted mainly by decreasing exposure due to heavy overgrowing with high macrophytes. The obtained data based on a visual census during relevant season, daytime and weather suggests that average density of edible frogs in beaver ponds was markedly higher than that in the comparable part of non-flooded fragments of small rivers: such a difference comprised 1.21 versus 0.15 individuals per 10 m2; or on average 1.76 and 1.02 individuals per 10 m2 of beaver ponds versus 0.10 and 0.21 individuals per 10 m2 of small watercourses.

This slideshow requires JavaScript.

A very good example of positive influence of beaver inundation on otter and mink was the results of our long-term monitoring of the beaver expansion in Volka small river catchment in Naliboki Forest. The catchment extended into an area of 500 k m2 (31 by 16 km) had about 380 kilometers of small watercourses such as small rivers and main forest drainage canals (excluding shallow canals). There, the increased numbers of beavers have substantially improved habitats for otters and mink. In effect of the habitat improvements numbers of otters along the small watercourses markedly increased. During the winter of 1982-1983, 19 beaver ponds and 22 otter locations (hereafter, an otter location means the particular place inhabited by either lone otter or family group of otters) were recorded. In the winter of 1997-1998, during the similar survey, 187 beaver ponds and 52 otter locations were found, and in the 2002-2003 winter, the populations increased again with 447 beaver ponds and 57 otter locations being recorded. So, during twenty years a three fold increase in the otter numbers was reported, which coincided with a 24 fold increase in the development of damming by beavers. Similar results were obtained in the Volka catchment in relation to the American mink.  It was less extended along a given watercourse (small river or canal), but overall the increase  in American mink numbers was much higher (approximately 10-20 fold) because of presence of plenty of secondary canals. Such secondary canals were almost not inhabited  by otters, while American minks populated nearly all of them, particularly, those inhabited by beavers. Explaining the revealed growth in the otter and American mink numbers in Volka catchment in connection with the beaver expansion recorded,  the obtained data on the changes in aquatic prey biomass and diets of semiaquatic mustelids may be considered.

Having data on aquatic prey biomass in beaver ponds in Volka catchment, we estimated that in 2006-2007 compared to the situation in 1982-1986 creation of numerous ponds by beavers led to great increase in aquatic prey biomass in all model small rivers: Volka, WS – 2.1 fold; Izliedz’, WS – 14.2 fold; Sivichanka, WS – 329 fold; Pruzhanitsa, WS – 779 fold; Zhawtsianka , WS – 924 fold; Volka, CS – 3.2 fold; Izliedz, CS – 5.0 fold; Sivichanka, CS – 38.5 fold; Pruzhanitsa, CS – 44.3 fold; Zhawtsianka, CS – 68.7 fold.

In the conditions of the considerable increase in aquatic prey biomass, the diets of both the otter and American mink changed markedly. Frequency of occurrences of fish in the species diets grew in both canalized small rivers and small rivers with natural riverbed in the warm and cold seasons. If before the beaver expansion otters and American minks mainly took dace, perch, mud loach, pike, roach, stone loach and three-spined stickleback (more or less in order of the prey  importance), then afterwards the predators species mainly preyed on pike, roach and mud loach, while perch, three-spined stickleback and stone loach became rare in their  catches. The consumption of amphibians by the semiaquatic mustelids declined there.