VC Soni Roosting is a typical bird behaviour where a group of individuals congregate in an area for a few hours Department of Bioscience, effected by an environmental signals and return to the same site with the reappearance of these signals. Surashtra University, Rajkot, Gujarat, India Present study was planned to know the selection criteria for roosting sites in House crows (Corvus splendens). To test the roosting site characteristics seventeen parameters were assumed which had Bhupat Radadia covered the broad aspects such as roost site characteristics, land use around sampling sites and Shri. M and N Virani Science anthropogenic pressure. Roost trees significantly tended to be taller by 25.39% more in tree height (t College, Rajkot, Gujarat, India Stat=3.0182>t Crit, 0.01>P value) and 36.45% more in canopy height (t Stat=5.470>t Crit, 0.01>P-value) compare to non-roost trees. While comparing the distance from feeding sites with non-roost sites, the roost sites were observed to be selected near to feeding sites as nearest feeding sites were 1379.5m significantly nearer than non-roost sites (t Test=3.619>t Crit, 0.01>P value). The distance of nearest tree (average distance of trees from four direction) from the roost sites was 33.34±26m which was 100.67 m nearer than non-roost sites (t=4.356>t crit, 0.01> P-value). Larger trees with greater canopy, nearby human habitation which provide them shelter and safety along with anthropogenic feeding opportunities and moderate vegetation patches near the roosting places were the characteristics preferred for roosting
purpose by house crows.
Keywords: Roosting Sites, Junagadh, Rajkot and House crow (Corvus splendens).
Introduction Roosting is a typical animal behaviour where a group of individuals congregate in an area for a few hours effected by an environmental signals and return to the same site with the [6, 13] reappearance of these signals (Finkbeiner et al. 2012, Richner et al. 1996) . Roosting behaviour is also observed in bats, primates, and insects but it is most common among birds (Beauchamp and Guy 1999) [1] which could be in thousands to millions of individuals in one roost (Pérez-García 2012) [12]. Belonging to corvidae family, house crows (Corvus splendens) are known to form larger communal roosts which contain a few hundred to over thousands of individuals (Ian 1977, Coombs 1961) [8, 2]. A study in Singapore found that house crows are
very common near human habitations and often busy streets and preferred to roost in area with [9] human activities and presence of human settlements (Kelvin et al. 2002) . They also prefer to roost close to feeding opportunities and in taller trees with dense crowns which is enclosed by tall-buildings (Kelvin et al. 2002) [9]. The present study was intended to identify the roosting site selection characteristics for the house crows (Corvus splendens). The survey was done by
studying roosting site of house crows at Junagadh and Rajkot cities of Gujarat.
Methodology To test the roosting site selection criteria seventeen parameters listed in Table 1 were assumed which had covered the broad aspects such as roost site characteristics, land use around
sampling sites and anthropogenic pressure. From total twelve roost sites, seven regular roost sites (3 roost sites from Rajkot and 4 roost sites from Junagadh) were selected to test the parameters. Further these parameters were compared with non-roost sites which were selected randomly in the study area which had the structure of cluster of trees same as roost sites.
Roost site characteristics Correspondence Tree measurements i.e. tree height, GBH, canopy height, canopy width, canopy density were Shabanam Saiyad measured and numbers of trees occupied were counted. If a roost site comprised ≥ 2 roost Shri. M and N Virani Science trees, average of all tree measurements was taken to be consider. Distance between roosting College, Rajkot, Gujarat, India trees forming one roosting site was measured. ~ 10 ~ International Journal of Fauna and Biological Studies
The distance from the base of the selected tree trunk to the Result nearest adjacent tree (average distance of trees from four Roost site selection direction), nearest road and nearest street lamp. Distance from Roost trees significantly tended to be taller by 25.39% more the base of the trunk to surrounding buildings was the mean in tree height (t Stat=3.0182>t Crit, 0.01>P value) and distance to the nearest buildings in 4 cardinal directions. The 36.45% more in canopy height (t Stat=5.470>t Crit, 0.01>P- distance of nearest building from tree trunk as well as the area value) compare to non-roost trees (Fig. 1.1). Further roost (length × width) of building was calculated that faced the sites had larger DBH by 69.38 % (Fig. 1.1) and denser crowns roost tree. The average distance of three nearest feeding by 17.9% more than non-roost sites. However significant opportunity from roost trees was measured. difference was not observed. Thus the greater DBH and higher canopy density were not required for roosting as crows Land use around roost sites were also observed to roost in moderate canopy density. Because selection of roost trees could be based on land use surrounding a site, the percentage of area covered by different types of land use around roost and noon-roost trees was determined. For assessment, all roost trees and non-roost trees were placed on google earth map and circle with a 1000 m radius was drawn, centred on the position of each tree. From the detailed map of google earth the percentage of different types of land use was estimated within the circle. A concrete environment was considered as built area. Open space consisted of cleared land, roads, play grounds and waste lands. Vegetation comprised areas covered by unmanaged vegetation, parks and over grown wastelands. Water included
areas of rivers, lake, ponds and canals. Fig 1: Tree measurements of roost and non-roost site Human-activity indices A human-activity index (Gorenzel and Salmon 1992) was Canopy cover was 91.54% significantly larger than of non- composed of multiple components, each given a score based roost sites (t=5.145>t crit, 0.01>P-value). Therefore by on potential human activity near roosting birds; the higher the comparing the roost site characteristic with non-roost sites; score (range=1 to 480), the greater the potential human the tree height, canopy cover and canopy height were found to activity. Components consisted of distance to the nearest be selection standards for roosting among all tree building (1=>30m, 2=> 2=30m, 3=>15-20m, 4=>7-15m, 5=0- characteristics. The distance of nearest tree (average distance 7m); distance to nearest railroad track (1=>150m, 2=0-150m); of trees from four direction) from the roost sites was night time traffic on the nearest road 1=0-10 vehicles per 10 33.34±26m which was 100.67 m nearer than non-roost sites m, 2=> 10 pedestrians per 10 m). The number of vehicles on (t=4.356>t crit, 0.01> P-value) (Fig. 1.2). This states that they the nearest road, number of pedestrian were obtained through prefer to roost in area nearby the high vegetation ten min surveys conducted at 2000. Individual components of comparatively. human-activity index were rated for all roost and non-roost Out of 12 roost sites, 8 roost sites were observed to roost near sites. The human-activity index was derived by multiplying to active roads and street lamps. The distance of roost site the scores of each individual component. from nearest road and street lamp were nearer by 33.18m (t Stat=5.590>t crit, 0.01> P value) and 581.55m (t Stat=6.953>t Table 1: House crow roost site characteristics of seven house crow crit, 0.01> P-value) respectively than non-roost sites (Fig. roost sites and non-roost sites. 1.2). Average distance to surrounding building was 47.29 m
Non-roost near than the non-roost sites (Fig. 1.2). No. Roost site Characteristics Roost Sites Sites 1 Tree height 18.67±9m 13.2±21m 2 Canopy height 11.8±10m 9.4±21m 3 Diameter at breast height 2.45±5.2m 1.12±3.2m 4 Canopy density 91.5±10 73.6±21 5 Canopy cover 98.4±14m 43.54±3.5m 6 Distance from nearest tree 33.3±12m 67.5±34.2m 7 Distance from nearest road 49.1±28m 82.3±54m Distance from nearest feeding 1673.55±213. 8 293.9±128m site 2m Distance from nearest street 9 30.7±10m 612.3±32m lamp 10 Distance between roost trees 7.45±3m 25.65±10.2m Fig 2: Distance of various anthropogenic features from roost and Average distance to 11 62.1±9m 109.4±2.3m non-roost sites. surrounding building 12 Dimension of nearest building 1342.5±28m2 967±103m2 The building that faced roost trees also tended to have greater 13 Built environment 65±32% 45±21% 14 Open space 22.7±12% 30±24% area than non-roost trees by 38.83%. Thus they prefer to roost 15 Vegetation 18.4±16.2% 10±11.2 near existence of human activates. Human-activity index was 16 Water 1.8±1.4% 5.9±11 found to be 56.26 % significantly higher (t Test=4.739>t crit, 17 Human Activity index 71.4±34 45.7±86 0.05> P value) than non-roost sites (Fig. 1.3). ~ 11 ~ International Journal of Fauna and Biological Studies
which provide them shelter and safety along with anthropogenic feeding opportunities and moderate vegetation patches near the roosting places were the characteristics preferred for roosting purpose by house crows. Following conditions were particular for roost site selection by house crows.
Roosting near to availability of food The primary requirement of the roosting birds is availability of plenty suitable roost trees in the proximity to suitable foraging sites. Studies on European starlings (Sturnus vulgaris), common grackles (Quiscalus quiscula) and American robins (Turdus migratorius) have revealed that roost sites were selected on the basis of their proximity to Fig 3: Human activity indices of roost and non-roost sites. good foraging areas. Common ravens (Corvus corax) were
observed to roost within 1 km of human-related food sources While comparing the distance from feeding sites with non- (Engle et al. 1992) [5]. roost sites, the roost sites were observed to be selected near to Crows were observed to prefer roosting in areas where ample feeding sites as nearest feeding sites were 1379.5m of food was available that can reduce commuting flight costs. significantly nearer than non-roost sites (t Test=3.619>t Crit, Thus distance from foraging grounds from the roost (Vyas 0.01>P value). 1996) as well as propinquity to food sources was a very Most stable roost sites were observed to roost near the human important factor in roost-site selection (Elsere 1984) [4]. habitats which was enclosed by settlements which may
provide them protection from predation as well as receiving Selection of trees with larger crowns and greater heights advantage to acquire warmness during low temperature. The for roosting widespread and repeated plantation of tall trees was a specific The crows used a verity of trees which had greater height and characteristic of roost trees. The distance between the roost larger canopy cover for roosting in the study area. Selection trees was from 7.45±3 meter and it was 18.11 meters lesser of large and dense crowns for roosting was because it may than non-roost sites. Therefore the distance between roost greatly reduce wind velocity and thus may provide protection trees was also very important factor for selecting roost sites. against convection and heat loss (Walsberg 1986) [15]. These These clusters of trees with less distance make merged crows trees provide protection from ground predator with its slender of all trees and may be helpful to effectively scan for potential and tall trunk which was difficult to climb. Very few predators and which also provide uniform and compact predatory threads to the crows at roost sites were recorded vertical arrangement of perch sites. during the study period with the presence of domestic dog Compare to non-roost sites, areas around roost sites had (Canis familiars), domestic cat (Felis catus) and raptors found significantly 28.30% more built area (t Stat=8.338>t Crit, to roost around the roosts. By roosting in trees with greater 0.01>P-value) (Fig. 1.4). Vegetation cover was found to be trunk height, the crows achieved greater distance from the 97% more around roost site compare to non-roost sites (Fig. pedestrians below. However the selection of tall roosts against 1.4) which was also significantly high (t Stat=5.922>t Crit, wind exposure appeared as a probable compromise between 0.01> P-value) than non-roost site (Table 5.17). Open space possible anti-predator advantages and energetics around roost site was 17% lesser than non-roost site (Fig. disadvantages (Draulans and Vessem 1986) [3]. 1.4).
Distance between roost trees The distance between roost trees was also very important factor. The average distance between roost trees was 7.47 metre. Well-spaced roost trees but not more than 10 metre between in urban areas may provide protection, since crows may be able to effectively scan for potential predators such as human, snakes and cats. The widespread and repeated plantation of tall trees was most favourable roosting habitat. Live and unbroken canopy cover of the preferred trees provided uniform and compact vertical arrangement of perch sites. This suggests the foliage preference by the crows. It
helps the birds in minimizing loss of the clear skies (Morse Fig 4: Area percentage of land use categories around Roost and 1980) [10]. Roosting of the crows in several patches even Non-roost sites within a roost sites was observed due to the availability of discrete patches of vegetation. Further, patchy vegetation at However no significant difference was found. Thus crows the roost sites probably provides clues to locate the roost from were observed to roost in high anthropoid habitat with the belt a distance. of vegetation. Roosting near to human places Discussion Urban areas may offer protection against native predators Roost site selection (Elserer 1984) [4] and therefore predator defence also may be Larger trees with greater canopy, nearby human habitation ~ 12 ~ International Journal of Fauna and Biological Studies
one of the explanations for crows roosting in highly urbanised and less vegetated areas. Most stable roosting sites may be achieved by selecting trees that more closely surrounded by human settlements, possibly provided protection from predation. All the roost sites except two in study area were found to be ensuing at very anthropoid habitation. This was also because house crows roosting in urban areas appear to be more tolerant to people (Wee 1989) [16]. Other benefit was also that they got maximum amount of food from various anthropogenic opportunities within the cities, therefore not going far and roosting where the food was available which reducing their cost of flight.
Reference 1. Beauchamp S, Guy A. The evolution of communal roosting in birds: origin and secondary losses. Behavioural Ecology. 1999; 10:675-687. 2. Coombs CJ. Rookeries and roosts of the rook and jackdaw in South-West Cornwall. Bird Study. 1961; 8(2):55-70. 3. Draulans D, Vessem VJ. The Adaptive sign of colonial breeding success. Ornis Scandinavica, 1986; 17:356-362. 4. Elserer LA. Communal roosting in birds. Bird Behavior, 1984; 5:61-80. 5. Engle AK, Young LS, Steenhof K, Roppe LA, Kochert MN. Communal roosting of common ravens in South western Idaho. Wilson Bulletin, 1992. 104(1): 105-121. 6. Finkbeiner W, Susan D, Adriana D, Briscoe H, Robert D. The benefit of being a social butterfly: communal roosting deters predation. Proceedings of the Royal Society of London, Biological Sciences, 2012; 279:2769- 2776. 7. Gorenzel WP, Salmon TP. Urban crow roosts in California. Vertebrate Pest Conference, 1992; 15:97-102. 8. Ian R. The social and spatial organization of winter communal roosting in Rooks (Corvus frugilegus). Journal of Zoology, 1977; 182:509-528. 9. Kelvin SH, Peh H, Navjot S. Characteristics of Nocturnal Roosts of House Crows in Singapore. The Journal of Wildlife Management. 2002; 66(4):1128-1133. 10. Morse DH. Behavioural mechanisms in ecology. Cambridge, Massachusetters, 1980; 8:67-69. 11. Morrison DW, Caccamise DF. Ephemeral roosts and stable patches. A radiotelemetry study of communally roosting starlings, Auk, USA, 1985; 102:793-804. 12. Pérez-García L, Juan P. The use of digital photography in censuses of large concentrations of passerines: the case of a winter starling roost-site. Revista Catalana d’Ornitologia, 2012; 8:123-129. 13. Richner J, Heinz O, Phillip Y. Communal life: Honest signalling and the recruitment center hypothesis. Behavioral Ecology, 1996; 7:115-118. 14. Vyas SN. Studies on some aspects of ecology of the Indian Black Ibis, Pseudibis papillosa (Temminck). At Rajkot. Ph.D. Thesis, Saurashtra University, Rajkot, India. 1996. 15. Walsberg GE. Thermal consequences of roost site selection: the relative importance of three modes of heat conservation. Auk, 1986; 103:1-7. 16. Wee YC. A guide to the wayside trees of Singapore. Singapore Science Centre, Singapore, 1989, 89-90.