© British Cave Research Association 2019
, Vol.46, No.3, (2019), 121–126
The world’s largest known subterranean fish:
a discovery in Meghalaya (NE India) of a cave-adapted fish
related to the Golden Mahseer, Tor putitora (Hamilton 1822)
Dan HARRIES 1, Thomas ARBENZ 2, Neelesh DAHANUKAR 3, Rajeev RAGHAVAN 4,
Mark TRINGHAM 5, Duwaki RANGAD 6 and Graham PROUDLOVE 7
1 Grampian Speleological Group, Edinburgh, UK.
2 Caving in the Abode of Clouds Project, Emetstrasse 34, 4713 Matzendorf, Switzerland.
3 Department of Biology, Indian Institute of Science Education and Research, Pune, India.
4 Department of Fisheries Resource Management, Kerala University of Fisheries and Ocean Studies,
5 Fellow of the Geological Society of London, Gloucester Speleological Society, University of Bristol
6 Department of Zoology, St. Edmund’s College, Laitumkhrah, Shillong 793003, India.
7 Department of Entomology, The Manchester Museum, University of Manchester, Manchester,
M13 9PL, UK.
Abstract: In February 2019 a troglomorphic fish was discovered in a cave in Meghalaya in northeastern India.
The largest individual seen in the cave was in excess of 400mm in standard length making it, by far, the largest
known subterranean fish found to date. Initial investigations indicate it is a close anatomical match to
but differs in its depigmentation, lack of eyes and in its subterranean habitat.
Keywords: Subterranean fish, cave fish, Meghalaya, standard length.
: 12 July 2019; : 23 September 2019.
By September 2019 there were 250 known species of
subterranean fishes (Proudlove, 2019). These animals live in
environments that are often extremely nutrient-limited because
of the absence of light and primary production. Consequently
most species of subterranean fishes are of relatively small size,
in order to survive on limited food resources. Figure 1 shows the
distribution of standard length for 195 of the 250 species. It is
clear that most species cluster in the range 20–130mm, with a far
smaller group between 130–230mm.
Two species exceed 300mm but both are eel-like and are very
thin in proportion to their length (195 species, mean 85.5mm,
range 23–420mm, data from Proudlove, 2019).
It has always been assumed that cave fishes exceeding
350mm would be most unlikely on resource grounds but this has
now been shown to be spectacularly wrong. The fish discovered
in Meghalaya in February 2019 is not only substantially longer
than the longest previously known species but is considerably
more bulky with a body mass likely to exceed that of the next
largest cave fish by at least an order of magnitude.
, Vol.46(3), 121–126, 2019
to Meghalayan cave exploration
The “” consists of a long-
running series of annual expeditions that have been exploring
and mapping Meghalayan caves continuously since 1992
(Arbenz, 2012, 2016). The project includes cave explorers and
scientists from around the world and works in association with
the Shillong-based “ ”. In
addition to exploration, mapping, and scientific monitoring and
analyses, they have also carried out a series of assessments of the
cave fauna (summarized by Harries , 2008), a systematic
inventory of cave biota that is still on-going. The project has
resulted in the discovery of a number of new cavernicolous
species in the past. These include a Huntsman Spider,
, (Jager 2005), two bats, and
, (Ruedi ., 2012) and a troglobitic fish,
(Kottelat ., 2007).
Large pale cyprinid fish have been noted in Meghalayan caves
on numerous past occasions (Harries , 2008). Such fish
have proved difficult to capture or to examine at close quarters.
With one notable exception all past records of closely examined
fish were found to be pale in colour but to have large normally
developed eyes. On the basis of this, it was assumed that they
were essentially epigean species present in the caves either as
accidental strays or as refugees displaced by the falling water
levels that occur during the dry winter months. The exception
was an observation made (Simon Brooks, pers. comm.) in
another Meghalayan cave in 1998. On this occasion several
large carp-like fish were captured and a number of these were
noted to lack eyes.
One of the team members who saw the fish in 1998 has also
seen the fish discovered in 2019 and maintains that they are
indistinguishable. Unfortunately, no clear photographs were
taken in 1998 and the specimens were not retained, so it is
not possible to verify the record independently. The 1998 site
is 8km to the southwest of the 2019 site. Current knowledge
of the hydrology and geology of the area would suggest that a
direct subterranean connection between the two sites is highly
unlikely, although they fall within the same overall surface river
, Vol.46(3), 121–126, 2019
Background and site description
In 2019 the expedition team were searching for caves in a remote
and densely forested area of the Jaintia Hills, Meghalaya. Several
caves were mapped in this area including the one that contained the
fish. Although the cave entrance is well known to the local people,
access to the inner parts of the cave requires specialist vertical
caving techniques. The exploration of the cave revealed a sequence
of entrance shafts that descend vertically for more than 100m.
Below the vertical entrance there is an extensive series of large
horizontal passages with numerous pools and streamways (Fig.2).
It is clear that the cave floods dramatically during the rainy months,
because patches of forest vegetation deposited by flood waters
were seen deep within the cave system. The new fish were noted
and photographed by a cave survey team on 18 February 2019.
On 19 February the site was revisited for further observations, and
a large specimen was captured and photographed before being
returned alive. A further visit was made on 21 February and a
medium-sized fish was collected for laboratory examination.
The entrance lies in a large, seasonally dry, rocky streambed
among forest, and comprises a large open pitch-head, beyond
which the entrance series is predominantly vertical with some short
(<20m) horizontal or steeply sloping sections. After descending for
~100m the entrance series drops into a horizontal streamway leading
to a large boulder-floored passage. The cave floor of the streamway
has some pools of standing water. It is predominantly rocky with
areas of bedrock, boulders and coarse gravel. This streamway is
relatively narrow (3–4m) but opens out into a considerably larger
boulder-floored passage. The floor of the boulder passage is mostly
elevated well above water level, although there are pools in places
along the western wall and in lower floor sections. Flood debris
consisting of forest vegetation is strewn along the floor indicating
that seasonally this area of the cave is flooded.
The fish were first encountered in a pool (~3 x 4m) that
spanned the passage width of the streamway about 70m
beyond the base of the entrance pitches and just before the
boulder passage. At this site two small (~15cm) and one
medium-sized (~25cm) individuals were seen. Considerably
more fish were present in pools below the western wall of
the boulder passage about 115m beyond the base of the
entrance pitches. The first of the pools was extensive (>10m
x 10m) but shallow (<1m), and occupied the entire floor of
the main downstream passage leading off from the boulder
chamber. Several fish were seen in this pool and appeared
to be predominantly small- or medium-sized individuals.
The second pool was a short distance (~20m) beyond the first,
and was accessed via a relatively low arch in the west wall
of the boulder passage. It was less extensive in area (~5m x
5m) than the first pool but was deeper (>1m) and appeared to
continue into a broad sump below the water surface. This pool
contained numerous (15–20) fish ranging from small (~15cm)
to large (>30cm) individuals. Numerous additional fish were
seen in several other pools (e.g. Figure 3) within the cave at
distances up to almost 500m from the base of the entrance
pitches. Based on these observations it seems likely that the
population must at least number in the hundreds.
Fish ranged in size from relatively small-sized (~10–15cm),
through medium-sized (~25–30cm) to large-sized (~35cm)
individuals. Although clearly of the same species, there are
distinct morphological differences between the small and the
large individuals. The smaller individuals are proportionately
more slender in body form than the relatively heavily built
larger individuals. But the degree of eye development is the
most striking morphological difference. In all cases the eyes are
regressed but in small individuals the eyes are clearly visible as
large dark patches under the skin surface (Fig.4).
, Vol.46(3), 121–126, 2019
In medium-sized individuals the eyes remain clearly visible but are
relatively small and less distinct than those of the small individuals.
In the largest individuals the eyes appear almost entirely regressed
and are not discernible without very close inspection (Fig.5). In
addition to the record length the biggest of the fishes are very bulky
(Fig.6) and it is assumed that they have a large and regular food
supply, with surface vegetation entering the cave via the entrance
shaft in rainy seasons. Despite this very large size for a subterranean
fish, these animals are clearly miniaturized compared to epigean
, which can reach 275cm and even in overfished areas
adults can reach up to 150cm (Bhatt and Pandit, 2016).
When first encountered the fish appeared unresponsive to light,
although they did react to the water disturbance created by cavers
wading through the pools. However, where water disturbance was
minimised they became inquisitive and appeared to be searching
actively for food. They were attracted to minor ripples caused by
patting the water surface and gnawed at boots and other items
placed in the water. The larger individuals seemed the most bold
and persistent in their search for food (which was the reason that
the large specimen in Figure 6 could be captured by hand despite
the lack of a suitable net).
Although it seems that they were initially unresponsive to light
they are certainly able to perceive light. When the deep pool off the
boulder passage was first examined, groups of several fish would
move into the shallow water (apparently searching for food) and
showed no reaction to caving torches shone in their direction at
The capture of the large specimen caused considerable
disturbance and after this event the fish in the pool became more
wary. In an effort to get more photographs of the fish, biscuit crumbs
were sprinkled in the pool to attract them into shallow water. After
waiting 15 to 20 minutes no fish had come close. We then left the
pool in order to examine other areas of the cave, and returned some
time later. When the torch was shone through the rock arch a group
of 10 – 15 fish could be seen feeding on the crumbs. When the
light was shone in their direction they dispersed immediately into
the sump leading from the pool before we were within 3m of the
pool edge. This pattern was repeated on two further occasions. It
seems that they can detect light but as a stimulus it initially lacked
meaning for them. The attack on them (in the form of the collection
of the large specimen) created an immediate association between
light and disturbance and thereafter they fled from the light.
It is well documented that some fish are capable
of learning rapidly to avoid stimuli associated
with an unpleasant event (such as capture)
and this avoidance behaviour can persist for a
considerable time. It seemed that the smaller
fish (with less regressed eyes) showed a more
rapid response to light and, even before they
experienced disturbance, they were more
difficult to approach than was the case for the
larger fish. It is perhaps remarkable that although
only one fish was attacked (captured) all the
other fish immediately learned to associate light
with danger and thereafter retreated when the
pool was illuminated. It might merely be that
they were reacting to the physical disturbance,
but many of these fish were not in the immediate
vicinity when the capture took place. It is
possible that chemicals released due to stress or
injury of the captured fish also played a role and
that such chemicals might be regarded as stress
pheromones (e.g. Smith, 1992).
This might provide an explanation as to how the
light-avoidance response appeared to be learned
by all individuals in the pool rather than just by
the very few that were in the vicinity at the time
of capture. See Pitcher (1993) and Godin (1997)
and references therein for information on learned
behaviour and predator avoidance in fish.
The fish is certainly by far the largest known
subterranean fish in the world (Fig.1). Initial
investigations demonstrate a close anatomical
match to . It differs, however, in a
lack of pigmentation, a lack of eyes and in its
The morphological features that correspond to
are as follows. It is a large carp with
big scales (Fig.7). The lateral line is complete
(Fig.7b) with 28 scales (Fig.7a). Lateral transverse
scales between dorsal fin origin to lateral line 3½
and between pelvic fin origin to lateral line 2½
(Fig.7a, b). Fleshy lips continuous at the angles
of the mouth with a continuous labial groove and
fleshy median mental lobe (Fig.8). Two pairs of
large barbels (Fig.8). Dorsal fin origin midway
between tip of snout and base of caudal, dorsal
spine bony, strong and smooth, slightly shorter
than depth of body (Fig.7). Dorsal fin (Fig.9 a, b)
with three simple and 9 branched rays (D iii 9).
Pectoral fin (Fig.9c, d, e) with one simple and 14
branched rays (P i 14). Ventral fin (Fig.9f) with
one simple and 8 branched rays (V i 8). Anal fin
(Fig.9g, h) with two simple and 5 branched rays
(A ii 5). The main taxonomic literature resources
used were Hora (1939), Sen and Jayaram (1982)
8b 8d 8e 8f
is distributed widely across all of the Indian
subcontinental region including India, Afghanistan, Pakistan,
Nepal, Bhutan and Myanmar (Lal, 1995; Bhatt and Pandit,
2016; Pinder , 2019). Significantly it is known from
the Garo Hills of Meghalaya (Dasgupta, 1982, 1991, a, b,
1993). It is therefore quite possible for the cave fish to have
evolved from this widespread epigean species. It is known to
inhabit rapid-flowing streams and pools such as those present
at lower altitudes in the Jaintia Hills (Joshia ., 2018).
It is also reported to be an opportunistic omnivorous feeder,
which would prove an adaptive advantage in the nutrient-
poor cave habitats.
Although the new fish is morphologically and meristically
indistinguishable from it is highly likely
that genetically they are somewhat different from this
species as a result of isolated evolution in the cave. That
evolutionary processes have taken place is evident from the
troglomorphic nature of the large fishes, which lack eyes
and melanin pigmentation. A small number of subterranean
fishes have retained the name of a previously known surface
fish despite being troglomorphic. The best studied of these
is Fowler and Steinitz 1956 in Oman
(Kruckenhauser , 2011; Kirchner , 2017) and these
studies might provide valuable models to follow in studying
the present subterranean fish.
Like many cave animals, the ability of the fish to move into
new areas tends to be constrained by the layout of the caves in
which they live. In some cases the entire world population of
a species might be restricted to a single cave system. If that is
the case with this fish, it would make the population extremely
vulnerable even to tightly localized impacts that might be
associated with changing land use, pollution or harvesting.
An overview of the local geology and hydrology is
informative in providing an indication of the potential
habitat extent of this fish population in terms of connectivity
to neighbouring cave systems and to areas of surface water
capable of supporting epigean populations of .
The entrance of the 2019 collection site is at the northern
upstream end of a tributary feeding into a network of
persistent surface rivers that flow southwards from the area
towards Bangladesh. This tributary has ephemeral surface
flow only during the wet season, but a subterranean hydraulic
connection is likely with other caves a few kilometres farther
south, which have large permanent springs feeding into the
southern river system. In the upstream direction subterranean
hydraulic connection to a neighbouring valley several
kilometres to the north is also possible. In this northern valley
some cave-stream sinks are known to drain southwards. Such
connections to catchment areas to the north of the known
site are also plausible, with the current flat-lying watershed
sensitive to changes in flow direction related to the active
tectonics and shifting tilt of the region.
So, in summary, subterranean connections to areas to the north
are possible but are insufficiently documented. Clearer potential
connections lie to the south, where surface rivers could allow
passage of fish moving upstream to enter cave resurgences. The
potential for such connections is likely to be far greater during
the wet months when increased flow will provide more surface
streams and more flooded subterranean passages that could act
as conduits for the dispersal of the fish populations.
The fish discovered in the Jaintia Hills of Meghalaya in
February 2019 is by far the largest troglobiotic fish yet known,
and is nearly 5 times the mean length (85mm) for all known
subterranean fishes to date. The only other species exceeding
300mm in length are eel-like Synbranchidae with nothing like
the bulk of the new fish. The large size of the latter is probably
related to a plentiful food supply.
Cave mapping Robin Sheen (Ireland), Anja Keatley (UK),
Marcel Dikstra (Netherlands), Jos Burgers (Netherlands), Dan
Harries (UK), Peter Ludwig (Austria), Uros Aksamovic (Serbia),
Laura Appleby (UK), Tim O’Connell (Ireland).
Photos Marcel Dikstra, Uros Aksamovic, Dan Harries.
Caving in the Abode of Clouds Project
Simon Brooks (UK),
Thomas Arbenz (Switzerland), Brian Kharpran Daly (Meghalaya).
Arbenz, T (Ed.), 2012. (Volume 1) Pala
Range and Kopili Valley. 72–105. [Matzendorf, Switzerland.]
Arbenz, T (Ed.), 2016. (Volume 2)
North Shnongrim Ridge and the Liat Prah Cave System. 102–109.
Bhatt, J P and Pandit, M K, 2016. Endangered Golden mahseer
Hamilton: a review of natural history.
, Vol.26, 25–38.
Dasgupta, M, 1982.
. PhD Thesis, North-Eastern
Hill University, Shillong, India.
Dasgupta, M, 1991a. Biometry of the mahseer, (Hamilton)
collected from Garo Hills, Meghalaya. ,
Dasgupta, M, 1991b. Food and feeding habits of the mahseer,
(Hamilton). , Vol.38, 212–217.
Dasgupta, M, 1993. A study on food and feeding habits of the mahseer
, Vol.3, 22–23.
Godin, J-G J, 1997. . [Oxford:
Oxford University Press.]
Harries, D B, Ware, J F, Fischer, C W, Biswas, J and Kharpran-Daly,
B D, 2008. A review of the biospeleology of Meghalaya, India.
, Vol.70, 163–176.
Hora, S L, 1939. The game ﬁshes of India III: the mahseer or the large scaled
barbels of India. 1. The putitor mahseer Hamilton
, Vol.41, 272–285.
Jager, P, 2005. New large-sized cave-dwelling Heteropoda species
from Asia, with notes on their relationships (Araneae: Sparassidae:
Heteropodinae): , Vol.112, 87–114.
Joshia, K D, Das, S C S, Pathak, R K, Khan, A, Sarkar, U K and Roy,
K, 2018. Pattern of reproductive biology of the endangered golden
mahseer (Hamilton, 1822) with special reference to
regional climate change implications on breeding phenology from
lesser Himalayan region, India. ,
Vol.46, 1289–1295. DOI: 10.1080/09712119.2018.1497493 .
Kirchner, S, Sattmann, H, Haring, E, Plan, L, Victor, R and
L, 2017. Investigating gene flow between the blind
cavefish and its conspecific surface populations.
, Vol.7(1), Article number 5130. DOI:10.1038/
Kottelat, M, Harries, D B and Proudlove, G S, 2007.
, a new species of cave loach from Meghalaya, India
(Teleostei: Balitoridae). , Vol.1393, 35–44.
Kruckenhauser, L, Haring, E, Seemann, R and Sattmann, H, 2011.
Genetic differentiation between cave and surface dwelling populations
of (Cyprinidae) in Oman.
,Vol.11, Article number 172.
Lal, M B, 1995
. [New Delhi: Hindustan Publishing
Menon, A G K, 1992. Taxonomy of the Mahseer ﬁshes of genus
Gray with description of new species from Deccan.
, Vol.89, 210–228.
Pinder, A C, Britton, J R, Harrison, A J, Nautiyal, P, Bower, S D,
Cooke, S J, Lockett, S, Everard, M, Katwate, U, Ranjeet, K, Walton,
S, Danylchuk, A J, Dahanukar, N and Raghavan, R, 2019. Mahseer
(spp.) fishes of the world: status, challenges and opportunities for
conservation. , Vol.29, 417–
Pitcher, T J, 1993. . [London: Chapman
Proudlove, G S, 2019.
http://cavefishes.org.uk read on 28 April 2019.
Ruedi, M, Biswas, J and Csorba, G, 2012. Bats from the wet: two
new species of tube-nosed bats (Chiroptera: Vespertilionidae) from
Meghalaya, India. Vol.119, 111–135.
Sen, T K and Jayaram, K C, 1982. Mahseer ﬁshes of India: a review.
, Vol.39, 1–34.
Smith, R J F, 1992. Alarm signals in fishes.
, Vol.2, Issue 1, 33–63.