Content uploaded by Pankaj Koparde
Author content
All content in this area was uploaded by Pankaj Koparde on Jun 07, 2015
Content may be subject to copyright.
Habitat correlates of Odonata in northern Western Ghats, India
21
Odonatologica 44(1/2) 2015: 21-43
1
st
June 2015
Habitat correlates of Odonata species diversity
in the northern Western Ghats, India
Pankaj Koparde
1,2
*, Prachi Mhaske
1
& Ankur Patwardhan
1
1
Department of Biodiversity, Abasaheb Garware College, Karve road,
Pune-411004, Maharashtra, India
2
Current address: Sálim Ali Centre for Ornithology & Natural History, Anaikatty
(Post), Coimbatore-641108, Tamil Nadu, India;
*
<pankajkoparde@gmail.com>
Received 10
th
May 2014; revised and accepted 22
nd
February 2015
Abstract. Sixty-two localities from Sahyadri Tiger Reserve, Maharashtra State, India, were
surveyed for habitat correlates of Odonata diversity. Proximate habitat variables (canopy cov-
er, area of water spread on transect, and altitude) and broad scale environmental variables
derived from climate database were used. Seventy species were recorded during the sur-
vey. Vestalis apicalis was found to be the most abundant species. Multiple regression analysis
failed to resolve relationship among variables. Proximate habitat variables, except altitude,
showed slightly higher contribution in shaping species richness and diversity than broad-
scale habitat variables. Canonical correspondence analysis based on species abundance data
and multiple variables suggested that canopy cover and area of water on the transect are driv-
ing species assemblages. Almost all of the Western Ghats endemics recorded during the sur-
vey were found to be associated with high canopy forests and streams, suggesting the critical
habitat requirement of these species. e study provides baseline and local habitat associa-
tion data on Odonata, which can be used as evidence in the conservation of the Sahyadri
Tiger Reserve corridor which is under threat of forest felling.
Key words. Dragony, Maharashtra, canopy cover, endemic, biodiversity, CCA, poisson
multiple regression
Introducon
Odonata, being one of the most widely studied insect groups, has been well-
documented in terms of its habitat requirements. Most Odonata are habitat
sensitive (e.g., S 2005; S et al. 2008). Presence
and/or absence of some species may indicate habitat health in terms of mi-
cro-habitat structure (e.g., O S 1996; S S
1996; S et al. 2008; G-A N-G
2010) and to some extent macro-habitat structure (S et al. 2003;
P. Koparde, P. Mhaske & A. Patwardhan
22
Odonatologica 44(1/2) 2015: 21-43
O 2005; C et al. 2010; D et al. 2011). Odonata species
assemblages might vary depending upon habitat quality (D L-
2003; S et al. 2003; G-A N-G
2010). Physical habitats are oen dynamic and it is hard to depict them in
quantitative terms, as not all the environmental variables can be measured.
Although much descriptive literature on habitat/s of Odonata is available,
few studies have conducted quantitative analyses of it. Such studies are im-
portant in detecting eect of habitat loss on Odonata. Few such studies have
been used in conservation practices. Given this, it is highly important to
determine local Odonata assemblage patterns and their habitat correlates.
e Western Ghats, a global biodiversity hotspot (M et al. 2000; M-
et al. 2011), have a rich Odonata fauna that is relatively well-
studied (F 1933, 1934, 1936; E R 2000;
E et al. 2005; S 2005, 2007; S S-
2005; R et al. 2010; K et al. 2012;
K S 2013; S et al. 2013; K et
al. 2014; R N 2014). A review by S (2007)
has reported 176 species from Western Ghats, of which 68 are known to
be endemic to the region. Despite this fact, the literature lacks articles on
quantitative habitat association of Indian Odonata species, apart from a few
research papers by S et al. (2005), S S-
(2005), D A (2007), S et
al. (2008), and K S (2013). e Western Ghats of
Maharashtra State are characterized by highly fragmented forest cover. e
fragmentation is increased by anthropogenic pressure (G et al. 2011).
Studies pertaining to Odonata of Western Ghats of Maharashtra are scarce
except for a few articles by K et al. (2012), K S-
(2013) and K et al. (2014). Species richness, diversity, and
composition are likely to change in response to changes in environmental
and habitat variables (e.g., S S 2005; F S 2008).
Most of the species endemic to Western Ghats are associated with closed
forests and owing water system (S et al. 2011). e habitat
association of these species has been described by many (e.g., F 1933,
1934, 1936; S 2005; K R 2013), in descriptive terms.
Previous studies on stream insects in Western Ghats by S
Habitat correlates of Odonata in northern Western Ghats, India
23
Odonatologica 44(1/2) 2015: 21-43
S (2005) suggest that genus-level richness is aected by
altitude, micro-habitat richness, canopy cover, number of dry months, and
annual rainfall in various habitats such as cascades, ries, and pools. In
another study by S et al. (2005) on similar lines, only micro-
habitat richness, canopy cover, and depth of water were found to have low
correlation with genera richness. In a highly fragmented landscape, such
as Western Ghats of Maharashtra, ne scale habitat variables (e.g., type of
aquatic system, canopy cover, and elevation) may aect Odonata diversity
more than broad-scale habitat variables (e.g., annual temperature and pre-
cipitation at the locality). Here, we hypothesize that these broad habitat vari-
ables may play a less important role in Odonata diversity than the aforemen-
tioned proximate habitat variables concerned with micro-habitat. In order
to understand the eect of dierent environmental and habitat variables on
Odonata diversity and assemblages, a study was carried out during 2011–
2013. is paper investigates how the scale of habitat correlates inuences
Odonata diversity and species composition. We also explore if the descrip-
tive literature on habitat association of endemic Odonata of Western Ghats
can be tested in a systematic framework.
Methods
Study site
Although highly fragmented, the landscape of the northern Western Ghats
retains some of the unique topographic features such as lateritic plateaux.
e part of the northern Western Ghats that lies in Maharashtra State, India,
is located roughly between 15°30’N to 20°30’N and 73°E to 74°E. Most of
this landscape is under legal protection. e largest protected area is Sahy-
adri Tiger Reserve (Fig.1), a 1,165 km²
forested landscape that lies between
16°57’N to 17°49’N and 73°35’E to 73°54’E. e core area of the Sahyadri
Tiger Reserve (STR) is 600.12 km² and the buer zone is 565.45 km². It is
spread across four districts, namely Satara, Sangli, Kolhapur, and Ratnagiri.
e forest type of the core area is primarily evergreen, semi-evergreen, and
moist deciduous forest; whereas that of the buer area is primarily dry de-
ciduous, and scrubland. Koyna wildlife sanctuary and Chandoli national
park are protected areas within the Reserve, situated on the banks of Shivsa-
gar and Vasantsagar water reservoirs respectively. e Reserve also supports
P. Koparde, P. Mhaske & A. Patwardhan24
Odonatologica 44(1/2) 2015: 21-43
many inland freshwater habitats suitable for Odonata including streams,
ponds, puddles, and marshes. e bioclimate of the region is not uniform
given the topography of the landscape. e vegetation structure changes
from evergreen, moist-deciduous forest patches in the core area to dry-
deciduous, scrub-forest in the buer area. e annual temperature ranges
from 13°C to 35°C. Annual mean precipitation is 178 mm. e study area
experiences three distinct seasons: Summer (February–May; mean temper-
ature 26°C, mean precipitation 24 mm); Monsoon (June–October; mean
temperature 22°C, mean precipitation 497 mm), and Winter (November–
January; mean temperature 22°C, mean precipitation 16 mm) (H et
al. 2005). e post-monsoon season (November–March) is the best season
to observe Odonata in the eld, in India (S 2005; N 2011).
Figure 1.-
veyed in northern Western Ghats, Maharashtra State, India (www.diva-gis.org/
Table 1.
Habitat correlates of Odonata in northern Western Ghats, India 25
Odonatologica 44(1/2) 2015: 21-43
Sampling and data collection
Sampling was carried out during the post-monsoon season (September–
March), when Odonata activity is at its peak. Belt transects of 500 × 10 m
were used to survey Odonata. An 8×40 (magnication × lens diameter) bin-
ocular was used to observe Odonata. Records other than on transects were
noted separately. Odonata were identied using eld-guides (S
2005; N 2011) and taxonomic monographs (F 1933, 1934, 1936).
e recorded species were listed according to the classication by D
et al. (2013) and S (2014). Area of water spread and canopy
cover on transect were measured on a 0–10 scale, 0 being absence. Area of
water spread on transect was estimated by observing the proportion of area
of water spread with respect to total area of transect. e position of transect
diered for dierent wetland systems. For instance, for streams and marshes,
the transect passed through the stream; whereas for ponds and reservoirs,
the transect was placed on the shore. Canopy cover was measured at inter-
vals of 50 m, on a 500 m transect, giving 11 survey points. e mode value
of canopy cover was used for further analysis. Transects with more than
30 % (more than 3 on 0–10 scale) of area of water spread were considered to
be wetland systems, which were further classied as pond/reservoir/stream.
Transects with average canopy cover of 40 % or more (equal or greater than
4 on 0–10 scale) were considered as closed forest systems. We created new
variables by combining canopy cover and wetland systems. Each locality
was dened based on these variables. ese new variables are representa-
tive of forest-wetland systems. Table1 summarizes the forest-wetland sys-
tems. In the present study, there were no areas in which we could nd pond/
reservoir in high canopy forest. Sixty-two localities from STR were studied
for Odonata diversity and habitat correlates (Tab.2). Sampling sites were
selected randomly covering the entire STR landscape and major wetland
systems. e data on mean annual precipitation, mean summer precipita-
tion, mean monsoon precipitation, mean winter precipitation, mean annu-
al temperature, mean summer temperature, mean monsoon temperature,
mean winter temperature, mean temperature of driest quarter, and mean
precipitation of driest quarter were obtained from WorldClim database at
30’ resolution (1 km² grid) (H et al. 2005). e data on altitude was
collected in the eld using a Garmin Map 60CSx GPS. For convenience, all
P. Koparde, P. Mhaske & A. Patwardhan
26
Odonatologica 44(1/2) 2015: 21-43
the variables have been divided into three sets: a) proximate habitat vari-
ables include canopy cover, area of water spread on transect and altitude;
b)broad-scale habitat variables were extracted from WorldClim database;
c) forest-wetland systems.
Data analysis
For all statistical analyses R soware v3.1.1 and PAST v3.2 (H et
al. 2001) were used. For Chao1 and Chao2 estimates of species richness,
EstimateS 9.1.0 (C 2013) was used. is measure provides the best
estimate of species richness (e.g., C C 1994; W
M 1998; W M 2001; C 2004). Simpson’s diver-
sity index, Fisher’s alpha index, relative abundance, and relative frequency
were calculated. Each species was assigned an occurrence value based on
percentage frequency. A Poisson multiple regression analysis was used to
determine drivers of diversity. For this, proximate and broad-scale habitat
variables were used. As variables derived from climate database are oen
correlated, a Kendall’s tau correlation analysis was carried out. Variables
that were highly correlated (τ > 0.8 or τ < -0.8 at p < 0.01) were discarded. We
Table 1.
FWS Abbre-
AWST Forest type Wetland
type
deciduous forest
NULL
Agriculture AGR NULL
NULL
forest
agriculture/scrub-land/
SAGR agriculture
canopy forest
Moist-deciduous/
Pond/reservoir in low
canopy forest
Agriculture/scrub-land/
pond/
reservoir
Habitat correlates of Odonata in northern Western Ghats, India
27
Odonatologica 44(1/2) 2015: 21-43
Table 2.
deciduous forest.
Locality Longitude FWS SF PA AWST
Adoshi 17.699 73.706 6 0
Akalpe 17.733 73.687 6 0
16.966 73.818 SAGR AGR 5 0
16.97457 73.801221 0 7
16.981 73.799 0 0
Arav 17.727 73.631 6 0
Atoli 17.30535 73.767214 0 8
17.30061 73.825293 3 9
17.581 73.864 AGR 3 0
17.546 73.843 5 1
17.16331 73.711052 5 8
16.927 73.864 5 0
Golivane 17.021 73.858 2 0
17.179 73.867 0 0
17.075 73.819 7 3
Kasani 17.19 73.896 7 0
Kathi 1 17.495 73.807 AGR 2 0
Kathi 2 17.493 73.814 5 0
Katrewadi 1 17.6348 73.8104 2 0
17.38966 73.657781 0 4
17.38339 73.666205 0 1
Kharoshi 17.77756 73.639083 0 0
Kolne 1 17.32225 73.721077 3 7
Kolne 2 17.329 73.73 3 0
17.397 73.675 AGR AGR 0 0
17.37593 73.706932 0 0
Male 1 17.31381 73.689008 0 6
Male 2 17.2886 73.7121 4 0
P. Koparde, P. Mhaske & A. Patwardhan
28
Odonatologica 44(1/2) 2015: 21-43
Locality Longitude FWS SF PA AWST
Mandur 1 17.14 73.88 AGR 7 3
Manoli 2 16.956 73.798 AGR 1 3
Manoli 16.944 73.803 5 2
17.67682 73.723697 3 4
Navja 17.42 73.741 5 2
Nechal 17.367 73.694 0 0
Paneri 17.222 73.828 0 1
Patharpunj 15.95104 74.0001 0 7
Pophali 1 17.44329 73.650413 AGR 2 0
Pophali 2 17.43074 73.667042 0 5
Pophali 3 17.42192 73.654726 5 0
Rundiv 17.24483 73.71603 2 7
Saatar 17.221 73.758623 5 0
Siddheshwar 17.19156 73.773 0 7
Sonpatra 1 17.43396 73.693199 0 6
Sonpatra 2 17.42641 73.67334 1 0
Tanali 17.128 73.8 2 0
Thoseghar 17.593 73.849 0 0
Thoseghar S1 17.574 73.807 1 0
Thoseghar S21 17.513 73.82 AGR 2 0
Thoseghar S22 17.512 73.816 2 0
Uchat 17.75652 73.631744 5 2
Udgiri 17.07479 73.836064 0 0
16.96138 73.848435 SAGR AGR 4 0
16.97756 73.846364 4 2
17.72765 73.588185 0 7
17.715 73.605325 1 3
17.4256 73.649428 4 1
17.439 73.847 AGR 2 0
17.55484 73.828 0 5
17.557 73.811 0 0
17.232 73.722 4 4
West Kusawade 17.5071 73.737405 3 2
17.57833 73.7459 0 7
Habitat correlates of Odonata in northern Western Ghats, India
29
Odonatologica 44(1/2) 2015: 21-43
ran twelve Poisson multiple regression models to check the eect of various
variables on Simpson’s diversity index and Fisher’s alpha index. ese mod-
els include proximate and broad-scale habitat variables in combination, as
well as in separate analyses. e signicance of each model was tested us-
ing a goodness of t test. If the test is statistically signicant, it suggests
that the data does not t the model. A canonical correspondence analysis
(CCA) was performed on species abundance data and all variable datasets.
A Monte-Carlo permutation test (n = 10,000) was performed on CCA for
signicance testing. A cluster analysis using the Bray-Curtis algorithm was
carried out with 10,000 bootstraps to understand overlap and uniqueness of
species between dierent forest-wetland systems in terms of species com-
position.
Results
Diversity and distribution of Odonata of Sahyadri Tiger Reserve
A total of 70 species of Odonata belonging to 45 genera and ten families
represented by 1,215 individuals was recorded during the survey. 64 species
were recorded on transects. 43 species belonged to the suborder Anisoptera,
and 27 species belonged to the suborder Zygoptera. Table 3 summarizes the
records. e Chao1 estimate of species richness ranged from 64–114 and
Chao2 ranged from 67–140. Vestalis apicalis Selys, 1873 was found to be
the most abundant species followed by Trithemis aurora (Burmeister, 1839)
and Pantala avescens (Fabricius, 1798) (Tab.3). Diplacodes trivialis (Ram-
bur, 1842) was the most widespread species in the study area followed by
P.aves cens and Trithemis festiva (Rambur, 1842). e most diverse locality
for Odonata fauna was Vakoli 1, followed by Uchat and Male 2.
Eect of habitat variables on Odonata diversity and species richness
In a preliminary correlation analysis on multiple variables, mean tempera-
ture and precipitation of summer, monsoon, winter, and driest quarter were
discarded. Altitude and mean annual temperature were highly negatively
correlated (τ = -0.96814, p = 0.00000001), therefore eect of altitude was
checked separately and all the analyses were re-run substituting altitude in
place of mean annual temperature. Table 4 summarizes output of models
1–12.
P. Koparde, P. Mhaske & A. Patwardhan
30
Odonatologica 44(1/2) 2015: 21-43
Table 3.-
-
No Taxon RA Oc Hcf Shcf Lcf Slcf Plcf Agr Sagr
1 Lestes elatus Hagen in Selys, 1862 0.0016 R
2 Lestes umbrinus Selys, 1891 0.0016 R
3 Protoscta hearseyi 0.0091 R
4 Vestalis apicalis Selys, 1873 0.0857
5 Vestalis gracilis 0.0016
6 Libellago lineata 0.0016 R
7 Heliocypha bisignata Hagen in Selys,
1853
0.0272
8 Euphaea fraseri 0.0124 R
9 Copera marginipes 0.0247
10 Copera viata Selys, 1863 0.0033 R
11 Caconeura ramburi 0.0058
12 Disparoneura quadrimaculata
0.0198
13 Elaoneura nigerrima (Laidlaw, 1917) 0.0041 R
14 Prodasineura vercalis (Selys, 1860) 0.0058
15 Aciagrion hisopa (Selys, 1876) 0.014
16 Aciagrion occidentale Laidlaw, 1919 0.0049 R
17 Aciagrion pallidum Selys, 1891 0.0025
18 Agriocnemis pygmaea 0.0066
19 Agriocnemis splendidissima Laidlaw,
1919
0.066
20 Ceriagrion coromandelianum
(Fabricius, 1798)
0.0008 R
Habitat correlates of Odonata in northern Western Ghats, India
31
Odonatologica 44(1/2) 2015: 21-43
No Taxon RA Oc Hcf Shcf Lcf Slcf Plcf Agr Sagr
21 Ceriagrion olivaceum Laidlaw, 1914 0.0041
22 Ischnura aurora 0.0016 R
23 Ischnura senegalensis
24 Pseudagrion decorum
25 Pseudagrion indicum 0.0016 R
26 Pseudagrion microcephalum
0.0041
27 Pseudagrion rubriceps Selys, 1876 0.0412
Aeshnidae Leach, 1815
28 Anax guatus 0.0008 R
29 Anax immaculifrons 0.0115
30 Gynacaha bayadera Selys, 1891 0.0066
31 Gynacaha dravida 0.0016 R
32 Gomphidia kodaguensis 0.0008 R
33 Heliogomphus promelas 0.0008 R
34 Icnogomphus rapax
35 Paragomphus lineatus (Selys, 1850) 0.0016 R
36 Epophthalmia viata 0.0016 R
37 Macromia 0.0025 R
Libellulidae Leach, 1815
38 Acisoma panorpoides 0.0016 R
39 Brachythemis contaminata
(Fabricius, 1793)
0.0033 R
40 Bradinopyga geminata 0.0107
41 Cralia lineata Förster, 1903 0.0091
42 Crocothemis servilia 0.0346
43 Diplacodes trivialis 0.0816
44 Hylaeothemis indica Fraser, 1946 0.0008 R
45 Idionyx spp. Hagen, 1867 0.0008 R
46 Indothemis carnaca (Fabricius, 1798) 0.0016 R
47 Lathrecista asiaca (Fabricius, 1798) 0.1402
48 Neurothemis fulvia 0.0157
49 Neurothemis intermedia
1842)
0.0239
50 Neurothemis tullia 0.0016 R
P. Koparde, P. Mhaske & A. Patwardhan
32
Odonatologica 44(1/2) 2015: 21-43
Eect of habitat variables on Odonata species assemblage
e CCA was statistically signicant based on Monte Carlo permutation
test (n = 10000) at p = 0.001 for the rst two axes. Axis 1 and axis 2 captured
31.48 % and 14.96 % of variation in data respectively. Area of water spread
on transect and canopy cover exerted strongest inuence on rst two axes
(Tab.5). Axis 1 was inuenced by canopy cover and axis 2 by area of water
spread on transects (Fig.2). e Bray-Curtis cluster analysis was supported
by a cophenetic correlation coecient of 0.9103. It showed that closed for-
est localities form a separate group in terms of species composition with
respect to other localities. Within low canopy forest areas, streams in low
canopy forest areas form a separate group (Fig.3).
No Taxon RA Oc Hcf Shcf Lcf Slcf Plcf Agr Sagr
51 Onychothemis testacea
52 Orthetrum chrysis
53 Orthetrum glaucum 0.0132
54 Orthetrum luzonicum 0.0107
55 Orthetrum pruinosum
1839)
0.042
56 Orthetrum sabina 0.0313
57 Orthetrum taeniolatum (Schneider,
1845)
0.0322
58 Orthetrum triangulare
59 Palpopleura sexmaculata (Fabricius,
1787)
0.0033
60 Pantala avescens (Fabricius, 1798) 0.0816
61 Potamarcha congener 0.0025
62 Rhodothemis rufa 0.0008 R
63 Tholymis llarga (Fabricius, 1798) 0.0041
64 Tramea basilaris
1807)
0.0049
65 Tramea limbata 0.0049
66 Trithemis aurora 0.0824
67 Trithemis fesva 0.0684
68 Trithemis kirbyi Selys, 1891 0.0074
69 Trithemis pallidinervis (Kirby, 1889) 0.0008 R
70 Zygonyx iris Kirby, 1869 0.0016 R
Habitat correlates of Odonata in northern Western Ghats, India
33
Odonatologica 44(1/2) 2015: 21-43
Table 4.
-
-
No Model
Goodness of
Model
Results
1
NO
2
ALT
NO
3 NO
4 NO
5
of any variables
6
of any variables
7
of any variables
8
of any variables
9 NO
10 NO
11 NO
12 NO
of any variables
Table 5. -
respondence analysis.
Axis 1 Axis 2
AWST Area of water spread on transect -0.31505 -0.46876
0.908706 -0.2412
ALT 0.063418 0.324567
PPT 0.199895 -0.08853
P. Koparde, P. Mhaske & A. Patwardhan
34
Odonatologica 44(1/2) 2015: 21-43
Axis 1 Axis 2
-0.08561 -0.31314
Pond/reservoir in low canopy forest -0.22914 0.031612
SAGR -0.13576 -0.19749
High canopy forest 0.771148 -0.12779
0.291415 -0.16173
-0.26073 -0.48257
AGR -0.23825 0.170657
Low canopy forest -0.25416 0.559464
AH Aciagrion hisopa -0.17007 -1.17183
AO Aciagrion occidentale -0.88573 -1.08728
AP Aciagrion pallidium 0.840023 -0.53094
APan Acisoma panorpoides -0.7317 -1.64806
Apyg Agriocnemis pygmya -0.81358 0.161933
AS Agriocnemis splendidissima -0.3866 -1.28271
AG Anax guatus -0.8209 2.06463
AI Anax immaculifrons -0.11968 -0.13052
Brachythemis contaminata -0.41514 1.87693
Bradinopyga geminata -0.12105 0.04665
Caconeura ramburi 2.90255 -0.19429
Ceriagrion coromandelianum -0.11607 -2.02898
Ceriagrion olivaceum 1.35895 -0.64487
Copera marginipes -0.51656 -1.55066
Copera viata 1.62109 -0.50493
Cralla lineata 1.71031 0.28603
Crocothemis servilia -0.36506 0.113149
Diplacodes trivialis -0.596 1.79608
Disparoneura quadrimaculata -0.54781 -0.91932
Elaoneura nigerrima -0.77748 -1.64685
Epophthalmia viata -0.47128 1.04956
Euphaea fraseri 2.20991 -0.47546
GK Gomphidia kodaguensis 0.319412 2.31644
Gynacantha bayadera 1.45126 0.034593
Gynacantha dravida 0.136866 -0.01031
HP Heliogomphus promelas 3.11169 -0.21512
HI Hylaeothemis indica 3.11169 -0.21512
Isp Idionyx sp. 3.11169 -0.21512
Habitat correlates of Odonata in northern Western Ghats, India
35
Odonatologica 44(1/2) 2015: 21-43
Axis 1 Axis 2
Indothemis carnaca -0.77581 1.4425
IA Ischnura aurora -0.76071 1.36924
LA Lathrecista asiaca 1.35767 0.243602
Lestes elatus -0.25383 3.21404
LU Lestes umbrinus -0.58309 -0.53566
LL Libellago lineata -0.42904 -0.03508
Msp Macromia sp. 2.13695 -0.408
NF Neurothemis fulvia 0.422692 -0.26085
NI Neurothemis intermedia 0.09702 -0.20195
NT Neurothemis tullia 0.154605 -1.6177
OG Orthetrum glaucum -0.05074 0.552532
OL Orthetrum luzonicum -0.68201 -0.47811
OP Orthetrum pruinosum -0.53654 0.245645
OS Orthetrum sabina -0.62998 0.286743
OT Orthetrum taeniolatum -0.59983 1.01873
PS Palpopleura sexmaculata -0.32919 -1.73163
PF Pantala avescens -0.5233 1.5364
PL Paragomphus lineatus -0.69852 -1.84859
Potamarcha congener -0.41571 0.913003
Prodasineura vercalis -0.62318 -1.45403
PH Protoscta hearseyi 2.84585 -0.26773
PI Pseudagrion indicum -0.7317 -1.64806
PM Pseudagrion microcephalum -0.20332 -1.21093
PR Pseudagrion rubriceps -0.67421 -1.20751
Rhinocypha bisignata 0.007045 -0.86936
RR Rhodothemis rufa -0.50573 2.84
TT Tholymis llarga 0.500541 -0.4299
Tramea basilaris 0.356902 1.18236
TL Tramea limbata -0.49017 0.972019
TA Trithemis aurora -0.63741 -0.47091
TF Trithemis fesva -0.56165 -0.60919
TK Trithemis kirbyi -0.76246 -1.42525
TP Trithemis pallidinervis -0.54835 2.68694
Vestalis apicalis 2.20657 0.249879
Vestalis gracilis 1.31079 -0.06045
ZI Zygonyx iris 1.26974 -0.30099
P. Koparde, P. Mhaske & A. Patwardhan
36
Odonatologica 44(1/2) 2015: 21-43
Discussion
e taxonomy and natural history of the Odonata of India is well known,
but few studies from India have focused on quantitative habitat correlates
of diversity and habitat associations. Species diversity and composition are
likely to change with respect to changes in micro-habitat and other environ-
mental variables. Detecting the relation between variables that might aect
species presence is crucial to understand gain or loss of species diversity
and to answer which species might get aected the most if environment or
habitat changes. is becomes important if species in question are endemic
or threatened. During this study, we were interested to know, given a set of
environmental variables, which variables might aect diversity and species
composition the most. All the localities in the study area were apparently
non-polluted, and are dicult to reach by roads. erefore, the inuence
of anthropogenic disturbance on all the sampling sites was either absent or
minimal. Most of the sampling was done in the post-monsoon season, when
Odonata activity is at its peak. e observed number of species, i.e., 70, falls
between the expected range of Chao1 (n = 64–114) and Chao2 (n = 67–
Figure 2.
-
Ghats. Text in red colour shows habitat variables.
Habitat correlates of Odonata in northern Western Ghats, India
37
Odonatologica 44(1/2) 2015: 21-43
Figure 3.
Reserve, northern Western Ghats, Maharashtra State, India. Nodal values show
140), suggesting that the study successfully recorded a near complete list of
Odonata in the study area. We recorded one species which was not previ-
ously known from Maharashtra State namely, Gomphidia kodaguensis Fra-
ser, 1923, which is known from central and southern Western Ghats (Goa,
Karnataka, and Kerala; R N 2014). A single specimen was
observed at Valvan. During the study we also obtained multiple records of
Elattoneura nigerrima (Laidlaw, 1917), Onychothemis testacea Laidlaw, 1902,
and Zygonyx iris Kirby, 1869, which are under-recorded from the northern
Western Ghats (F 1933; B et al. 2013; K et al. 2014).
e Poisson multiple regression analysis on species richness (models 1–4,
table 4) indicate eect of canopy cover, area of water spread on transect, and
annual mean precipitation, but not all the predicted models had a good t
to the data. Models based on Simpson’s diversity index (models 5–8, table 4)
were statistically signicant (models tted data), but did not show any eect
of variables on the diversity index. e regression analysis on Fisher’s alpha
index (models 9–12) showed that canopy cover and area of water spread on
transect are aecting diversity index values; however the models did not
P. Koparde, P. Mhaske & A. Patwardhan
38
Odonatologica 44(1/2) 2015: 21-43
have a good t to the data. To check the eect of broad-scale habitat varia-
bles, i.e., mean annual temperature and precipitation, and proximate habitat
variables, i.e., canopy cover, area of water spread on transect, and altitude,
we carried out the analyses in combination as well as separately. In all the
analyses, there was no statistical signicance relating the variables to diver-
sity index or species richness. Diversity and species richness indices sum-
marize abundance and species number data in a single value. As dierent
species may respond dierently to changes in environment, summarizing
them in a single variable may result in loss of data. is might be the rea-
son behind failure of the multiple regression analysis to identify any signi-
cant variables. e multiple regression analysis shows that all the proximate
habitat variables, except altitude, inuence species richness and diversity as
compared to broad-scale environmental variables in the study area. Howev-
er, in all these models we did not nd a case where the model tted the data
and the variables in the model showed statistically signicant eect on de-
pendent variables. S S (2005) and S-
et al. (2005) found that altitude, micro-habitat complexity, canopy
cover, number of dry months, and annual rainfall were driving diversity and
distribution of stream insect communities in Kudremukh national park,
Western Ghats. In their studies, mean annual temperature did not show a
signicant correlation with diversity and distribution of aquatic insects. In
the present study, we did not nd correlation between mean annual tem-
perature and diversity and species richness of Odonata, but could recover
weak eect of canopy cover. Further, we found a weak eect of area of water
spread on transect on diversity and species richness of Odonata. S-
S (2005) and S et al. (2005) looked
at aquatic insect communities and present study restricts to Odonata. is
might limit the power of comparison of results obtained.
To explore the eect of dierent habitat variables on species composition,
we performed canonical correspondence analysis. CCA revealed that can-
opy cover and area of water spread on transect are driving species com-
position, as compared to other variables (Fig.2; Tab.5). Proximate habi-
tat variables, except altitude, were found to be major driver for changes in
species assemblages. O (2005) reports that shading, water speed and
Habitat correlates of Odonata in northern Western Ghats, India
39
Odonatologica 44(1/2) 2015: 21-43
water permanence aected Odonata species assemblages in a study done
at Papua New Guinea. O (2005) measured shading in terms of percent
of shade on water body, which is analogous but not same as that of canopy
cover. Most of the Western Ghats endemic species recorded during the sur-
vey were found to be associated with high canopy cover, high canopy for-
ests, and streams in high canopy forests (Fig.2). Bray-Curtis cluster analysis
(Fig.3) also showed a similar trend with high canopy forests having very low
similarity in species composition as compared to other forest-wetland sys-
tems. e pattern of Western Ghats endemic species associated with dense
forests has been discussed by F (1933, 1934, 1936), S et
al. (2011), and K R (2013). Similar studies from Mt. Hamiguitan
wildlife sanctuary in the Philippines by V M (2010)
has reported that dense montane forests showed high number of species
endemic to Mindanao as compared to other forest types.
Although lot of descriptive literature on habitat association of Indian
Odonata is available, quantitative testing of the same has not been attempt-
ed before. In the present study, we found a similar habitat association of
species from Western Ghats of Maharashtra, as described in literature. Most
of endemic species recorded in the study were found to be associated with
high canopy cover and streams. According to S et al. (2011),
endemic Odonata of Western Ghats are mostly found in riverine habitats
such as montane streams and rivers. Western Ghats endemics like Proto-
sticta hearseyi Fraser, 1922 and Euphaea fraseri (Laidlaw, 1920) are known
to inhabit montane streams in dense forests (F 1933, 1934, 1936;
S 2005; N 2011). Heliogomphus promelas (Selys, 1873) and
Caconeura sp. occur in good water and forest quality (N 2011). Most
of the localities in which these species were recorded were either semi-ev-
ergreen or moist-deciduous forests. Species, other than those endemic to
Western Ghats, showed consensus in their habitat association as described
in literature and observed during the present study. For instance Lestes ela-
tus Hagen in Selys, 1862, is known to inhabit ponds or lakes (S-
2005; N 2011) in scrub jungles during the post-monsoon and drier
season (F 1933). Similarly, Indothemis carnatica (Fabricius, 1798) is
known to inhabit heavily weeded ponds and lakes. During our study, we
found both these species near ponds with low or no canopy cover (Fig.2).
P. Koparde, P. Mhaske & A. Patwardhan
40
Odonatologica 44(1/2) 2015: 21-43
In the present study we considered only three proximate habitat variables
viz. canopy cover, area of water spread on transect and altitude, which may
not represent overall eect of all the proximate habitat variables such as wa-
ter quality, water ow-rate, substrate structure, and vegetation structure in
combination, on species assemblage. Adding more continuous habitat vari-
ables to the analysis may be more informative on the eect of habitat change
on Odonata diversity.
e tiger corridor between Koyna wildlife sanctuary and Chandoli national
park is highly fragmented in terms of forest cover. Most of the new records
were observed in Bahe, Atoli, and Kolne 1 localities, all of which fall in the
corridor area. Other than these areas, Chandel and Rundiv locality of Chan-
doli national park (Fig.1) were found to be rich in endemic species. e
corridor area has smaller patches of forest, which are under high pressure of
wood-cutting by local villagers. Developmental activities such as wind-mills
and road widening were observed to be the causal agents of habitat distur-
bance. Studies have shown that anthropogenic disturbances can have detri-
mental eects on Odonata species (M 2009; V M-
2010) or may aect species assemblage (D A
2007; R MI 2009; K S 2013). S-
et al. (2011) points towards agricultural pollution and urban and
industrial development as the major threats to the Odonata fauna of West-
ern Ghats. Many endemic species are narrowly distributed across Western
Ghats, occurring in only small patches of suitable habitats (S
et al. 2011; K et al. 2014). To ensure conservation of such species,
it is highly important to protect their micro-habitats, especially closed for-
ested streams. e core areas of Sahyadri Tiger Reserve are under legal pro-
tection and with current status of the population of endemic species, it is
easy to conserve them. However, the conservation of the species in buer
areas, especially in the corridor area, is highly dependent on conservation
of forests within.
Acknowledgments
e present study was supported under the Indian Space Research Organ-
ization-University of Pune collaboration initiative. We thank Dr M.C. Ut-
Habitat correlates of Odonata in northern Western Ghats, India
41
Odonatologica 44(1/2) 2015: 21-43
tam, Director of ISRO-UoP for his encouragement. Dr. K.A. Subramanian
of ZSI, Kolkata, was always helpful in conrming the identication of the
specimens. We are thankful to him for reviewing the rst dra of manu-
script. We also thank forest department sta of Sahyadri Tiger Reserve for
their help. We also thank Principal, Abasaheb Garware College, Pune, for
his support. Field level support and assistance from post-graduate students
and research sta of Department of Biodiversity, Garware college, is grate-
fully acknowledged. We also thank the anonymous referees for their critical
comments.
References
Records
of the Zoological Survey of India 347: 1-60
odonate richness and diversity with eleva-
-
-
Insect Conservaon and
Diversity 3: 302-312
-
-
-
Philosophical Transacons
of the Royal Society of London
Sciences) 345: 101-118
2003. Odo-
upper Guinean forest. Archiv für Hydrobio
lo gie 157: 397-412
2013.
-
Zootaxa 3703 (1): 1-82
S. 2007. An-
Journal
of Insect Science 7: 1-9
rain forest as indicators of changes in bio-
Tropical Zoology 24: 1-9
Wildlife Sanctuary, Kerala, India. Records of
the Zoological Survey of India 98: 157-167
-
P. Koparde, P. Mhaske & A. Patwardhan
42
Odonatologica 44(1/2) 2015: 21-43
-
est lakes: rapid species turnover driven by
Insect Conservaon and
Diversity 1: 169-179
-
Taylor and Francis, London
-
Taylor and Francis, London
-
Taylor and Francis, London
-
2011. Report of the
Western Ghats ecology expert panel. Part I.
2010. Richness and structure of an Odonata
-
Odonatologica
39: 305-318
-
sion 1.89. University of Oslo, Oslo
-
global land areas. Internaonal Journal of
Climatology 25: 1965-1978
Sciences, Kerala
A.
Jour
nal of threatened Taxa 6: 5744-5754
A. K.A. 2013.
-
nata (Insecta) of Mula and Mutha river ba-
sins, Maharashtra, India. Journal of threat
ened Taxa 5: 4084-4095
2012. Insecta: Odonata. In:
Fauna of Maharashtra: State Fauna Series
20: 397-428. Zoological Survey of India,
Kolkata
-
isoptera). Odonatologica 38: 247-255
2011. Glo-
-
-
-
berg
-
-
Nature 403: 853-858
of Orissa and eastern India. Wildlife Organi-
rainforest in Papua New Guinea. Interna
onal Journal of Odonatology 8: 243-257
Habitat correlates of Odonata in northern Western Ghats, India
43
Odonatologica 44(1/2) 2015: 21-43
-
-
terns at new ponds in South Africa. Odona
tologica 25: 49-58
2014. Further ad-
Goa, India. Journal of threatened Taxa 6:
5585-5589
O.
of Goa. Journal of threatened Taxa 2: 805-
814
-
anthropogenic disturbance. Insect Conser
vaon and Diversity 2: 73-80
-
Bio
logical Conservaon 78: 279-288
2003.
-
ate approach. Hydrobiologia 497: 169-180
-
nates of the Western Ghats: Habitat dis-
K.A. 2014. A checklist of
Odonata (Insecta) of India. Zoological Sur-
vey of India, Kolkata
Ghats. Current Science 89: 976-987
-
Journal
of Insect Science 5: 1-10.
2008. Odonata as indicators of riparian
western Karnataka, India. Fraseria (NS) 7:
83-95
bio diversity in the Western Ghats, India:
2013. Idionyx
species. Zootaxa 3652: 277-288
-
diversity. Biodiversity & Conservaon 14:
165-178
-
versity and status of Odonata across veg-
Asian Journal of
Biodiversity 1: 25-35
2001. Species
Ibis 143:
413-419
-
Parasitology 116:
395-405
P. Koparde, P. Mhaske & A. Patwardhan
44
Odonatologica 44(1/2) 2015: 21-43Odonatologica 44(1/2) 2015: 44
