Akiyoshi-do Cave is a show cave located at 34°13.44'N, 131°18.14'E and roughly at 84 m above sea level (Figure 2). It is located in Akiyoshi-dai Plateau National Park, Yamaguchi Prefecture in Japan, and is known as one of the most visited show caves in the country due to its development, big stalactites and biodiversity of cave inhabitants. The cave is on the lists of Natural Monuments (since 1922), Special Natural Monuments (since 1955), Ramsar Convention Wetlands (since 2005) and Japan Geoparks (since 2015) and offers people learning experiences in a natural environment.

The Chimachida rimstone pools are one of the main tourist spots in Akiyoshi-do Cave located about 330 metres from the main entrance. They are composed of approximately 500 large and small pools arranged within a diameter of 20 m (Figure 3). The upper part is in contact with the tourist trail, has a gentle slope over the lower part and water eventually joins the cave river towards the main entrance, which is southwards. There is a wall with a film of flowing water on the opposite side across the tourist trail and the water is transported by a pipe under the tourist trail and discharged into the pools. In addition, pumped-up cave river water is continuously provided to the pools; this water supply started on 19 October 1971 with the aim of improving the water quality, in response to the fact that the pools’ contamination had become noticeable. Spring water at the bottom of some pools is another source of water supply to the pools. Dripping water from the ceiling is another supply source but much smaller compared with the others.

Figure 2. 

Location of the Akiyoshi-do Cave in Japan and of the surveyed pools inside the cave. The cave river flows to the south. The map of Akiyoshi-do Cave was drawn based on karusuto.com (https://akiyoshido.karusuto.com/html/guide/).

Figure 3. 

The main entrance of Akiyoshi-do Cave (left) and the Chimachida Rimstone Pools (right).

Ten pools were selected from the Chimachida rimstone pools and named CH1 to CH10. Visual observations were made of the colour of the pool bottom, the particle size of sediments, the presence of human-related debris and other features, for qualitative evaluation of each pool on 16 November 2015. The water quality of pools and the population densities of the two amphipod species were measured/observed on 17 November 2015, 17 February, 17 May and 8 August 2016.

Nakamura and Kuramoto (1978) classified the pools into three types, based on the sediment of the pool bottom and the presence/absence of the amphipods. Type A pools were defined by the deposition of reddish-brown clay which suggested little inflow of organic matter. Type B pools were characterised by the deposition of black-brown clay, suggesting a large influx of organic matter. Type C pools were characterised by a larger amount of polluted black-brown clay deposition compared with type B pools. In this study, pools were classified in the same way, plus, the water source was considered. The small differences between the previous and our classification concerned the pools unselected in the previous study, the improved condition of the pool bottom, thus reclassifying some from type C to B, and worsened condition of the pool bottom, thus reclassifying some from A to C.

The 10 study pools can be classified into three major types, according to the type of water supplied to the pools, either water on the walls across the tourist trail or pumped-up cave river water. The first type, named type A pool, are pools that receive a larger amount of water from the wall, this being transported under the tourist trail and supplied to the pool, and this corresponds to CH1-5. Another type, named type B pool, are pools which receive a large amount of pumped cave river water, and this corresponds to CH6-9. Finally, there is a type C pool characterized by the absence of water exchange, and it corresponds to CH10. In addition, CH10 has a higher elevation of the pool bottom than adjacent pools and is isolated by a rimstone wall thicker than 5 cm.

P. akatsukai is a troglobite, with white body, reduced eyes, developed sense of touch, and is adapted to an oligotrophic environment (Uéno 1927; Tomikawa and Nakano 2018) (Figure 4). Many of these amphipods live in places where the flow of water is slow, such as pools and underground rivers. G. nipponensis is a troglophile that lives both inside and outside of caves, prefers organic matter more than P. akatsukai and has no preference for water flow speed (Figure 4).

Figure 4. 

Pseudocrangonyx akatsukai (left) and Gammarus nipponensis (right).

A 50 cm square quadrat was created in order to measure the population densities of P. akatsukai and G. nipponensis. The quadrat was placed in the pools and the number of amphipods was counted with the naked eye.

Water temperature, pH and electric conductivity were measured on site using a HORIBA handy meter (model SSS054, D-54). The quantity of dissolved oxygen was measured using a HORIBA handy meter (model SS054, D-55s).

Water samples were collected on site using two 250 ml water bottles per site, placing them in a cool box at below 10 °C and transported to the laboratory.

The alkalinity (C) was determined by titration method. Chemical oxygen demand (COD) (mg/l) was determined by the potassium permanganate titration method. Total nitrogen (TN) was measured by Merck’s pack test. The total phosphorus content (TP) was also determined by Merck’s pack test.

The number of all bacteria in the water sample, which is known as one of the indicators of water pollution, was determined. A test tube containing 9 ml of sample water was prepared and 1/10 of a formalin solution 1 ml was added and mixed well to fix bacteria in the sample. Bacteria are stained by adding DAPI (4’, 6-diamidino-2-phenylindole) nucleic acid stain, water is filtered through an anopore inorganic membrane filter produced by Whatman plc and the total number of bacteria was counted using a fluorescence microscope.

The number of viable bacteria is also a biological index of water pollution, as well as the total number of bacteria. Undiluted sample water, 1/10 diluted sample water and 1/100 diluted sample water were placed in 1/10 Nutrient Broth, 1.5% agar medium and cultured for 3 days in an incubator set at 20 °C. Two plates were made for each dilution level. The experiment was performed on a clean bench environment. Bacterial colonies were marked on plates by dots with a pen until the cultivation period ended.

Canonical Correlation Analysis was performed with software R in order to clarify the correlation between each of the water quality parameter and the two amphipod species. Total nitrogen and total phosphorus as nutrients for the amphipods, and both total number of bacteria and total number of viable bacteria as indicators of water pollution, were used in the analysis.

The results of macroscopic inspection of the pools are shown in Figure 5 and Table 1.

Table 1 shows contaminants and other characteristics of the three types of pools. In the type A pools, no contaminants such as green algae or hair, were observed or there were in very small amount compared with the type B pools. By contrast, in the type B pools, contaminants, such as hair and plastic fragments, were particularly noticeable and abundant green algae had developed. On the other hand, in the type C pool, green algae development and contaminants, such as hair and wood chips, were more prominent than in type A and B pools.

Figure 5. 

Pools with arrows pointing the flowing directions. Water from the walls and drips was supplied to CH4 and water flowed to CH2 and CH1 via CH5. Pumped cave river water was supplied to CH8 and CH9 and water flowed to CH6 and CH7.

Table 2 shows the mean values of the water parameters across the Chimachida rimstone pools. The water temperature was slightly lower than 17 °C, the mean temperature of the atmosphere in Akiyoshi-do Cave. The values of total nitrogen and total phosphorus of type A, B, C pools in each survey are also represented in Table 2. The mean values of total phosphorus were in increasing order in type A, type B and type C pool, of 0.05 mg/l, 0.11 mg/l and 0.17 mg/l, respectively. The mean values of chemical oxygen demand were 0.2 mg/l for both type A and type B pools, and 0.6 mg/l for the type C pool (Table 2). The type C pool showed the total number of bacteria of 1.3×10⁶ Cells/ml and the total number of viable bacteria of 1.3×10⁴ CFU/ml, both exceeding the mean values of the type A and B pools. The mean values of the total number of bacteria were 5.7×10⁵ Cells/ml and 9.8×10⁵ Cells/ml for type A and B pools, respectively. The mean values of the total number of viable bacteria were 3.7×10³ Cells/ml and 1.1×10⁴ Cells/ml for type A and B pools, respectively.

Table 3 shows the population densities of P. akatsukai and G. nipponensis of all surveys in this study. The population densities of both species varied from 0 to 10 ind/m². The population densities of P. akatsukai were 0–8 ind/m², 0–6 ind/m², and 0–1 ind/m² for type A, B, and C pools, respectively. The population densities of G. nipponensis were 0–10 ind/m², 0–4 ind/m², and 0–20 ind/m² in type A, B, and C pools, respectively. Table 4 and Figure 6–8 show the comparison between population densities of 1970s and present. The maximum population densities of P. akatsukai in type A pools in the 1970s was 12 ind/m², and that of the present was 12 ind/m². The maximum population densities of G. nipponensis in type A pools in the 1970s and the present were both 8 ind/m². The maximum population densities of G. nipponensis decreased from the 1970s to the present in type B and C pools: from 260 to 20 ind/m², from 935 to 10 ind/m², respectively.

Figure 6. 

Fluctuation of the population densities of Pseudocrangonyx akatsukai and Gammarus nipponensis in the type A pools (1971–1975, 2015–2016).

Figure 7. 

Fluctuation of the population density of Gammarus nipponensis in the type B pools (1971–1975, 2015–2016).

Figure 8. 

Fluctuation of the population density of Gammarus nipponensis in the type C pools (1971–1975, 2015–2016).

Number of both P. akatsukai and G. nipponensis increased in autumn, but decreased in winter (Table 3). In type B pools P. akatsukai increased in spring and summer, and G. nipponensis increased in autumn and winter (Table 3). In the type C pool, P. akatsukai was not observed, whereas G. nipponensis was observed in summer and autumn (Table 3).

The analysis results are shown in Figure 9. P. akatsukai had a relatively strong negative correlation with total nitrogen. G. nipponensis had a positive correlation with the total phosphorus, the total number of bacteria and the total number of viable bacteria.

Figure 9. 

Result of CCA analysis on the amphipod abundances and water characteristics. TN: total nitrogen, TP: total phosphorus, TB: total number of bacteria, TVB: total number of viable bacteria. Other symbols mean pools surveyed in each season by “month-pool number”: for example, 5-CH1 corresponds to CH1 pool surveyed in May.

The obtained correlations corroborate with the findings reported by Nakamura and Kuramoto (1978) that P. akatsukai, a troglobite, does not like the nutrient-rich environment (total nitrogen) and, hence, it was suggested that the increase in organic matter contamination caused by tourism could reduce their habitat availability. On the other hand, the population density of G. nipponensis, a troglophile, increased as the number of microorganisms and the amount of nutrient (total phosphorous) increased. However, it was observed that the population density of G. nipponensis decreases as water pollution progresses (Nakamura and Kuramoto 1978). It, thus, might indicate a threshold for the concentration of organic matter in the water in which G. nipponensis can live.

The study of Nakamura and Kuramoto (1978) showed that the population density of P. akatsukai increased in winter and decreased in summer and that of G. nipponensis increased in summer and decreased in winter, in type A pools. The present study showed different results in type A and B pools.

It was stated that P. akatsukai did not inhabit pools CH7-9 in the 1970s. In the four surveys conducted in this study, P. akatsukai was confirmed in those three pools. Pumped river water supply started in 1971, possibly contributing to the habitat expansion of P. akatsukai. On the other hand, in CH10, the type C pool, formerly classified as a type A pool, P. akatsukai was not found in the present survey. The CH10 pool is surrounded by a thick wall of 5 cm or more and water exchange was not confirmed. No other water supply exists for CH10. This pool is located at the top of Chimachida rimstone pools and is one of the pools with the highest frequency of contact with tourists. Therefore, relatively large amounts of tourist-derived organic matter are supplied and eutrophication progresses. This can be the reason why P. akatsukai, which is known not to inhabit eutrophic environments, retreated from this pool.

The population density of both P. akatsukai and G. nipponensis in any type of pool decreased compared to the 1970s. In the absence of population density data prior to the 1970s makes it difficult to discuss the impact of tourism development in Akiyoshi-do Cave.

Akiyoshi-do Cave experienced the highest number of tourists in the 1970s and has shown a gradual decreasing trend since then to the present, but the population density of P. akatsukai and G. nipponensis became significantly smaller.

It is difficult to conclude that the source of organic matter contamination to the Chimachida rimstone pools is limited to cave tourism. The tourist activities on the surface where Akiyoshi-do Cave is developing, the so-called Akiyoshi-dai Plateau, also take part in environmental alteration of the water inside the cave. There is a museum, observation decks and shops located on the plateau within the range of the catchment area of Akiyoshi-do Cave. The sewage from these tourist facilities is transported to the basement through the underground sewer and treated. If sewage leaks out it can contaminate the Chimachida rimstone pools. In the 1990s, the sewer pipe was damaged, contaminating the upwelling water and creating obnoxious odours at another place not connected to the Chimachida rimstone pools in Akiyoshi-do Cave. Kuramoto and Nakamura (1995) and Kuramoto (1995) conducted a survey to find the influence the leaked sewage had on the fauna in Akiyoshi-do Cave. As a result, they found that many of the subterranean animals that were identified at the time of the survey by the Japan Cave Groundwater Research Association (1957) were observed, even in 1995. In addition, organic matter originating from animal manure can reach Akiyoshi-do Cave from poultry farms and an abandoned ranch in the catchment area of Akiyoshi-do Cave. Kawano and Fujii (1985), Yoshimura and Inokura (1992) and Haikawa (2006) reported the water system of Akiyoshi-dai Plateau National Park. Their tracer experiments gave no details on the inflows that reach the pools in the Akiyoshi-do Cave.