A comparison of canopy spiders in tropical forest in Hainan Island, China

, ,

(1. College of Life Science, Shenyang Normal University, Shenyang 110034, China;2. Institute of Zoology, Chinese Academy of Sciences, Beijing 100101, China)

Hainan Island, which includes the largest area of tropical forest in China, is considered to be part of the Indo-Burma biodiversity hotspot[1]. It represents only 0.35% of the area of China, but contains 82 species of mammals(17.4%), 344 species of birds(29.2%), 141 species of amphibians and reptiles(19.8%)[2]. However, the forests in Hainan Island were decreased quickly from 25.7% in 1956 to 7.2% in 1980, with average deforestation rate about two percent per year[2]. This may disrupt ecosystem functions and constitute a major threat to the long-term biodiversity conservation[3].

Forest canopy are among the most species-rich yet most highly threatened terrestrial habitats[4], and it remain one of the world’s least-known habitats[5]. Spiders are one of the most abundant group of predators in forest canopies[6]and also rich in species richness, e.g., 472 species were collected from a total of 1 000 m2of projected area in Xishuangbanna, SW China, including more than one hundred species reported as new to science recently[7]; 555 species were collected in tropical lowland forests of Peru[8]; 578 species in Dipterocarp lowland rain forests of Kinabalu National Park in Sabah, Malaysia on Borneo[9]. However, only limited experimental information exists for canopy-dwelling spiders, especially in tropical Asian[6-12]. Arthropod inventories in Hainan Island have mainly focused on ground or understory of forest[13-14,16], unfortunately, no previous studies have focused on canopy spider diversity in Hainan Island.

Although morphospecies were often be used as surrogates for species in studies such as environmental monitoring and biodiversity conservation[16], there are still inconvenience in ecological studies when used canopy spiders as surrogates. Because tree canopies in tropical forest usually harbored large number of spider species with a high proportion of new species[7], as well as familiar phenomenon of dimorphism in some spider families and a large proportion of juveniles, making it a daunting task to study canopy spider diversity in tropical forest. However, it will be feasible if used higher taxa surrogates at family and guild level. Higher trophic levels have been repeatedly found to be especially sensitive to environmental change[17]. For example, Shochat et al.[18]determined how land-use alteration influenced spider diversity and tested how habitat type and productivity affected spider diversity and abundance using the data of spider families. Tilman et al.[19]found that functional diversity is one of the most important parameters used to explain how ecosystems work and adapt to change. Studying functional groups can be useful to investigate assemblage response to climate change[17]and habitat disturbance[20].

In this study, we compared canopy spider assemblages among three nature reserves with different human management, elevation and ages. Our objectives were 1) investigated spider assemblages in response to human management(e.g., scenic spot and forest recovery), 2) compared spider assemblages across nature reserves.

1 Material and methods

1.1 Study site

Fig.1 Location of selected fogging stations in the present study in each nature reserve in Hainan Island, China

Hainan Island is located at the southern China with a total area of 33 920 km2. Lying at the northern edge of tropical Asia and in the tropical monsoon climate zone, the climate of Hainan Island is clearly divided into dry(November to April) and wet(May to October) seasons with an annual average precipitation of 1 500～2 600 mm, which mainly concentrated in June to November. Annual mean temperature is 22～26 ℃[21].

This study was conducted in tropical montane rain forest and tropical lowland rain forest of Diaoluoshan (18°40′～18°48′N, 109°45′～110°02′E), Jianfengling(18°23′～18°52′N, 108°44′～109°02′E) and Yinggeling(18°49′～19°08′N, 109°11′～109°34′E) Nature Reserves, situated in the southeast part, southwest part and south-central part of Hainan Island respectively (Fig.1). Natural vegetation where this study was conducted is classified as a tropical montane rain forest at regions above 800 m elevation; at regions below 800 m, is tropical lowland rain forest(Tab.1).

Tab.1 Environmental variables of the 24 sites investigated in Hainan Island, China

* But for local visiting for fungus.

1.2 Spider Sampling

Samples were collected from 6 to 25 August 2010. Twenty four samples were collected each comprised twenty sheets and approximate 300 m2in total size. A portable thermal fogging machine (Thermal Fogger TS-35A, China) was used to disperse insecticide from the ground and then the mist drifted up through the canopy. Fogging was conducted before sunrise to minimize fog scatter, since wind speeds are low at these times. Sampling never took place after rain, during windy or misty conditions.

The fogger operated for 20 minutes on each sample used 2-litre of a 2.2% solution of pyrethroid dissolved in diesel oil(Xingcheng Company, Guangdong, China), which with low mammalian toxicity. A 2 h period was allowed for spiders to drop onto sheets in order to maximize the collection and reduce the number of spiders escaping due to recovery from the toxic effect of the insecticide. All specimens were collected and then preserved in 75% alcohol.

Families of adult spiders were grouped into guilds according to their web-building and prey-catching behaviors[22-23]. 1)ambush hunters(AMH): Deinopidae, Selenopidae, Thomisidae, Oonopidae; 2)orb web weavers(ORW): Araneidae, Nephilidae, Symphytognathidae, Tetragnathidae, Uloboridae; 3)foliage hunters(FOH): Corinnidae, Gnaphosidae, Liocranidae, Clubionidae, Ctenidae, Miturgidae, Oxyopidae, Salticidae, Scytodidae, Sparassidae, Theridiosomatidae; 4)sensing web weavers(SEW): Hersiliidae; 5)sheet web weavers(SHW): Agelenidae, Linyphiidae, Ochyroceratidae, Pisauridae, Tetrablemmidae; 6)space web weavers(SPW): Dictynidae, Pholcidae, Theridiidae; 7)specialists(SP): Mimetidae, Telemidae, Zodariidae.

All adult specimens were identified to families. Due to the difficulty involved in identifying juveniles, they were excluded from the analyses. All specimens were stored in the Zoological Museum of China.

1.3 Statistical analyses

The number of families was estimated using the programme EstimateS, species richness estimators of Chao 1 was used[24]. The rarefaction curves were employed to compare the completeness of sampling for each region using EstimateS[24]. One-way ANOVA and Bonferroni correction comparisons were used to compare the relative abundance of various spider guilds among the three sites. Homogeneity of variances was tested using Levene’s test. Data were checked for normally distribution using Kolmogorov-Smirnov test. All data were calculated and analyzed with SPSS 15.0 computer program (SPSS Inc. USA).

The similarity among sampling sites was depicted as Bray-Curtis similarities[25]by using both families and guild compositions. Non-metric multidimensional scaling (NMDS) were employed to explore the relationship based on the similarity values. Analysis of similarities (ANOSIM) was performed between each pair of habitats to determine the significance level. NMDS construction and ANOSIM analyses were performed by PAST[26].

2 Result

2.1 Spider assemblage composition and abundances

A total of 45 632 specimens were collected in the present study, among them only 8 422(18.5%) were adults, representing 32 families. The mean portion of adults was highest in Yinggeling Nature Reserve(22.5%), followed by Diaoluoshan Nature Reserve(19.0%), and lowest in Jianfeng Nature Reserve(15.8%).

Family richness was similar among three nature reserves, 24 familes were found in Diaoluoshan and Yinggeling, and 25 families were found in Jianfengling. Four families were particular to Diaoluoshan and Jianfengling, respectively; three families were particular to Yinggeling(Tab.2). The composition of canopy spiders varied among the three nature reserves. Theridiidae(39.1% of total adult individuals) and Clubionidae(22.9%) are abundant families in Diaoluoshan, comprised 62.0% of all adult spiders. Theridiidae(37.7%), Clubionidae(16.3%) and Thomisidae(10.5%) are abundant families in Jianfengling, contributing approximately 64.5% of total adult individuals. Theridiidae(43.8%), Clubionidae(18.1%) and Salticidae(11.6%) are abundant families in Yinggeling, together comprised 73.5% of the adults. Specifically, Theridiidae is the most abundant family in almost every sample, but two sites in Jianfengling, and Ochyroceratidae is also found to be rich in abundance in canopies of Hainan Island, constituting 9.1%, 7.8% and 9.5% of all adults, in Diaoluoshan, Jianfengling and Yinggeling, respectively.

Tab.2 Total number of individuals of canopy spiders classified by family and guild sampled from Hainan Island, China

ContinuedtoTab.2

GuildDiaoluoshanindividuals%Jianfenglingindividuals%Yinggelingindividuals%Totalindividuals%Foliagehunters(FOH) Corinnidae20．180．3100．4200．2 Gnaphosidae000050．250．1 Liocranidae381．170．340．2490．6 Clubionidae81222．938316．346018．1165519．7 Ctenidae00100010 Miturgidae822．31054．5371．52242．7 Oxyopidae551．6662．8180．71391．7 Salticidae2888．11626．929411．67448．8 Scytodidae00140．600140．2 Sparassidae110．370．340．2220．3 Subtotal128836．47533283232．9287334．1Sensingwebweavers(SEW) Hersiliidae90．350．250．2190．2 Subtotal90．350．250．2190．2Sheetwebweavers(SHW) Agelenidae20．1000020 Linyphiidae912．6582．5532．12022．4 Ochyroceratidae3229．11827．82419．57458．8 Pisauridae10000010 Tetrablemmidae00100010 Subtotal41611．824110．329411．695111．3Spacewebweavers(SPW) Dictynidae0070．3140．6210．2 Pholcidae421．2451．9371．51241．5 Theridiidae138439．188337．7111143．8337840．1 Subtotal142640．393539．9116245．9352341．8Specialists(SP) Mimetidae501．4200．9230．9931．1 Telemidae00001010 Zodariidae50．1150．690．4290．3 Subtotal551．5351．5331．31231．4 Totalofindividuals3541234425378422 Totaloffamilies24252432

Fig.2 Spider diversity comparison among three regions in Hainan Island, using accumulation curves

The similar rare faction curves(Fig.2) were observed in the three regions suggested that canopy spider richness in term of family level differed slightly. Three curves all nearly reach an asymptote, indicating that the sampling was approach completeness. The expected numbers of spider family estimated by Chao 1(Chao 1984) were 24.2±0.6 in Diaoluoshan, 25.1±1.6 in Jianfengling, and 24.9±1.6 in Yinggeling. Coinciding with the above results, the expected numbers also indicated few families would be recorded by further sampling.

2.2 Spider guild structure

Fig.3 Relative abundance of spiders in tree canopies in Hainan Island (Numbers on the top of columns represent spider individuals collected in each region; percentages in the columns represent the ratio of dominant guilds.)

Spider guild composition in the Jianfengling differed from Diaoluoshan and Yinggeling in having a higher proportion of AMH. AMH was remarkable higher in Jianfengling as the third dominant guild, constituting 12.5% of all specimens. The majority of spiders were SPW, constituting39.9%～45.8% of all specimens; followed by FOH, constituting 32.1%～36.4% of all specimens. SHW web weavers was the third dominant guild in Diaoluoshan and Yinggeling, constituting 11.8% and 11.6% of all specimens; and SHW was the forth dominant guild in Jianfengling, constituting 10.3% of all specimens (Fig.3). Data about spider guild abundance were used to compare among regions using one way-ANOVA and LSD mean comparisons. Only the relative abundance of AMP was significantly higher in Jianfengling than others(F=6.47, df=23,P<0.01); other guilds showed no obvious different among regions (P>0.05).

2.3 Comparison of family and guild compositions between regions and other factors

The NMDS plots generated from relative abundance of different canopy spider families in sampling sites located in different nature reserves of Hainan Island were similar to that of guilds(Fig.4). Both of the NMDS plots showed that sampling sites from Diaoluoshan and Jianfengling were well separated from each other, but for one site that belongs to Jianfengling was more related to Diaoluoshan. The sampling sites in Yinggeling grouped loosely and showed some overlap with that of Jianfengling and Diaoluoshan. Pair-wise ANOSIM tests based on both family compositions and guild compositions also showed significant differences between Diaoluoshan and Jianfengling(P<0.01). Family compositions in Yinggeling did not differ significantly with Diaoluo shan and Jianfengling(P>0.05); guild compositions in Yinggeling showed significant difference with Diaoluo shan(P<0.01), whereas no significant difference with Jianfengling(P>0.05, Tab.3).

Fig.4 NMDS plots each study sites in Hainan Island, China generated by family composition (a) and guild composition (b) of canopy spiders(Triangle, Diaoluoshan; Square, Jianfengling; Solid circle, Yinggeling.

Tab.3 Results of pair-wise ANOSIM tests comparing spider family and guild compositions between different regions in Hainan Island, China

With respect to disturbance and the age of the forest, the results of NMDS plots generated for relative abundance of different spider families revealed no obvious patterns(P>0.05, Tab.3). Whereas, the guilds showed that both the disturbance and age of the forests have difference patterns(P<0.05, Tab.3), though sampling sites showed some overlap between disturbed/recovery forests with none-disturbed/old-growth forests(Fig.4).

3 Discussion

A related high number of canopy spiders (45 632 individuals) were collected in the present study, whereas, only a small proportion (18.4%) of which was identifiable adults. The ratio of adult spiders was similar to Zheng et al.[7]who collected 24 323 individual canopy spiders including 4 999 adults(20.6%). Whereas, it is much lower when compared to that of Sliva[8]who found 38% of the spiders were adults; and Russell-Smith and Stork[12]found 32% of the spiders were adults. A Chao 1 estimate to compare the observed family richness indicated that the collection was almost completed. However, spiders that are not readily collectable by canopy fogging method, such as cryptic species, do not take into account in the estimate.

The prevalent families of Araneae in canopies of Hainan Island were different from other studies conducted in montane rain forest in tropical Asia. Zheng et al.[7]found Theridiidae(21.9%), Oonopidae(18.2%) and Thomisidae(16.3%) were the dominant families in the secondary tropical montane rain forest in Menglun Natural Reserve(appr. 950 m), Yunnan Province, China. Pham[27]found that Theridiidae(37.2%, 37.4%), Araneidae(10.8%, 10.2%) and Salticidae(8.0%, 12.5%) in the primary/recovered tropical montane rain forest in Tam Dao National Park (appr. 1 000 m) in Vietnam. In another South-east Asian island of Sulawesi, Indonesia, Russell-Smith & Stork[11]reported that Theridiidae(30.0%), Salticidae(12.0%), Clubionidae(12.0%) and Oonopidae(10.0%) dominated the canopy spider fauna in montane rain forest (appr. 1 150 m). In contrast, Oonopidae and Araneidae only accounted for 1.4% and 2.9% of adult spiders in Hainan Island. It is worth noting Ochyroceratidae was very common in tree canopies of Hainan Island, it accounted for between 7.8% and 9.1% of adult spiders. Curiously, Ochyroceratidae was rare or absent in tropical forest canopies in previous studies[7,11-12,28-29].

The ridiidae almost always dominate tropical canopy spider communities in terms of individuals[6,9-12,28,30-32]. The previous results were consisted with the present study, in the all twenty four samples Theridiidae accounted for 40.1% of total adult individuals and was the most abundant family in twenty two sites. Clubionidae is often found to be a dominant family in tropical forest canopies[6,11-12]; it contributed 19.7% of total adult individuals and was the second dominant family in all the three regions. Salticidae contributed 11.6% in Yinggeling, while it was 8.1% and 6.9% in Diaoluoshan and Jianfengling. Thomisidae contributed 10.5% in Jianfengling, whereas it was only 4.8% and 2.7% in Diaoluoshan and Yinggeling.

It has been shown that the richness and abundance of canopy spiders vary according to microhabitat qualities[8]. Hainan Island was abundant in rainfall but was inequality in the distribution. The eastern part is windward side, rich in rainfall; the western part is leeside, less rain[21]. Diaoluoshan and Jianfengling are located on the opposite side of south-east and south-west of Hainan Island. As a result, spider family composition was distinguished significantly in family and guild levels according to the NMDS plots and pair-wise ANOSIM tests. This indicated that the distribution of canopy spider was strongly impacted by regional environment and climate. Yinggeling is lying in the central of southern part of Hainan Island and has nearly equal distance to Diaoluoshan and Jianfengling. Thus the spider composition in Yinggeling was similar to both Diaoluoshan and Jianfengling. Although many commonly found families were familiar in all the three regions, some prevalent families were more associated with one region and most rare families were also regionally distributed.

Zheng et al.[7]showed that anthropogenic disturbance depress vegetational complexity thus strongly influenced the assemblages of canopy spiders. Floren and Linsenmair[33]and Zheng et al.[34-35]also reported that anthropogenic disturbances reduced the size of the species pool dramatically. In the present study, we found that tourist sight-seeing did have effect to canopy spider guilds, whereas, it has little effect to canopy spider assemblages in term of family. In addition, similar to Floren et al.[36]and Purchart et al.[37], our result showed that the age of forests also affects spider guilds assemblages. While a rather long recovery time seemed narrowing the difference of canopy spider assemblages, especially in term of families.

4 Acknowledgements

Thanks go to Yucheng Lin, Jie Liu and Caixia Gao for help identification of spiders. The late Zuwei Zha, Dr. Daiping Wang, Hongliang Zhao and Zhiguo Hong kindly helped in the field work.

[ 1 ]MYERS N,MITTERMEIER R A,MITTERMEIER C G,et al. Biodiversity hotspots for conservation priorities[J]. Nature, 2000,403:853-858.

[ 2 ]CHEN Lingzhi. The biodiversity of China: State of the art and conservative strategies[M]. Beijing: Science Press (in Chinese), 1993:181-193.

[ 3 ]FOLEY J A,DEFRIES R,ASNER G P,et al. Global consequences of land use[J]. Science, 2005,309:570-574.

[ 4 ]STORK N E. Insect diversity: facts, fiction and speculation[J]. Biological Journal of the Linnean Society 1988,35:321-337.

[ 5 ]OZANNE C M P,ANHUF D,BOULTER S L,et al. Biodiversity meets the atmosphere: A global view of forest canopies[J]. Science, 2003,301:183-186.

[ 6 ]STORK N E. The composition of the arthropod fauna of Bornean lowland rain forest trees[J]. Journal of Tropical Ecology, 1991,7(2):161-180.

[ 7 ]ZHENG Guo,LI Shuqiang,YANG Xiaodong. Spider diversity in canopies of Xishuangbanna rainforest (China) indicates an alarming juggernaut effect of rubber plantations[J]. Forest Ecology & Management, 2015,338:200-207.

[ 8 ]SILVA D. Species composition and community structure of Peruvian rainforest spiders: a case study from a seasonally inundated forest along the Samiria river[C]∥Proceedings of theth International Congress of Arachnology. Geneva: Revue de Suisse Zoologie, 1996:597-610.

[ 9 ]FLOREN A,DEELEMAN-REINHOLD C. Diversity of arboreal spiders in primary and disturbed tropical forests[J]. Journal of Arachnology, 2005,33(2):323-333.

[10]H?FER H,BRESCOVIT A D,ADIS J,et al. The spider fauna of neotropical tree canopies in Central Amazonia: first results[J]. Studies on Neotropical Fauna and Environment, 1994,29:23-32.

[11]RUSSELL-SMITH A,STORK N E. Abundance and diversity of spiders from the canopy of tropical rainforests with particular reference to Sulawesi, Indonesia[J]. Journal of Tropical Ecology, 1994,10(4):545-558.

[12]RUSSELL-SMITH A,STORK N E. Composition of spider communities in the canopies of rainforest trees in Borneo[J]. Journal of Tropical Ecology, 1995,11(2):223-235.

[13]TANG Guo,LI Shuqiang. Crab spiders from Hainan Island, China (Araneae, Thomisidae)[J]. Zootaxa, 2010,2369:1-68.

[14]TONG Yanfeng,LI Shuqiang. The goblin spiders of the genus Opopaea (Araneae, Oonopidae) in Hainan Island, China[J]. Zootaxa, 2010,2912:23-43.

[15]REN Haiqing,CHEN Jian,YUAN Xingzhong,et al. Composition and diversity of spiders in the rubber plantation and natural forest of Hainan Island, China[J]. Acta Ecologica Sinica, 2016,36(6):1774-1781.

[16]OLIVER I,BEATTIE A J. Invertebrate morphospecies as surrogates for species: A case study[J]. Conservation Biology, 1996,10(1):99-109.

[17]VOIGT W,PERNER J,JONES T H. Using functional groups to investigate community response to environmental changes: two grassland case studies[J]. Global Change Biology, 2007,13(8):1710-1721.

[18]SHOCHAT E,STEFANOV W L,WHITEHOUSE M E A,et al. Urbanization and spider diversity: Influences of human modification of habitat structure and productivity[J]. Ecological Applications, 2004,14(1):268-280.

[19]TILMAN D,REICH P B,KNOPS J,et al. Diversity and productivity in a long-term grassland experiment[J]. Science, 2001,294:843-845.

[20]LAVOREL S,MCINTYRE S,LANDSBERG J,et al. Plant functional classifications: from general groups to specific groups based on response to disturbance[J]. Trends in Ecology & Evolution, 1997,12(12):474-478.

[21]ZHANG Liming,WEI Zhiyuan,QI Zhiping. Characteristics of rainfall and evaporation of different region in recent 30 years in Hainan Province[J]. Chinese Agricultural Science Bulletin, 2006,22:403-407.

[22]TSAI Z I,HUANG P S,TSO I M. Habitat management by aboriginals promotes high spider diversity on an Asian tropical island[J]. Ecography, 2006,29(1):84-94.

[23]CARDOSO P,PEKR S,JOCQUé R,et al.Global Patterns of Guild Composition and Functional Diversity of Spiders[J]. Plos One, 2011,6(6):e21710.

[24]COLWELL R K. EstimateS: Statistical estimation of species richness and shared species from samples[EB/OL]. 8.2 ed. Computer program and documentation[2017-12-08]. http:∥viceroy.eeb.uconn.edu/EstimateS.

[25]KERBS C J. Ecological methodology[M]. New York: Harper & Row, 1989.

[27]PHAM D S. Spider diversity in the tropical forest of Northern Vietnam in relation to regional condition and habitat structure[D]. Beijing: Chinese Academy of Sciences, 2009.

[28]GUILBERT E,CHAZEAU J,LARBOGNE L D. Canopy arthropod diversity of New Caledonian forests sampled by fogging: preliminary results[J]. Memoirs of the Queensland Museum, 1994,36:77-85.

[30]BASSET Y,ARTHINGTON A H. The arthropod community of an Australian rainforest tree: Abundance of component taxa, species richness and guild structure[J]. Australian Journal of Ecology, 1992,17(1):89-98.

[31]MAJER J D,RECHER H F,POSTLE A C. Comparison of arthropod species richness in Eastern and Western Australian canopies: a contribution to the species number debate[J]. Memoirs of the Queensland Museum, 1994,36(1):121-131.

[32]GUILBERT E. Arthropod biodiversity in the canopy of New Caledonian forests[M]. STORK N E, ADIS J, DIDHAM R K (Eds), Canopy arthropods. London: Chapman & Hall, 1997:265-277.

[33]FLOREN A,LINSENMAIR K E. The influence of anthropogenic disturbances on the structure of arboreal arthropod communities[J]. Plant Ecology, 2001,153:153-167.

[34]ZHENG Guo,LI Shuqiang,YANG Xiaodong. Biodiversity of ground-dwelling spider in six forest types in Xishuangbanna Daizu Autonomous District, SW China[J]. Acta Entomologica Sinica, 2009,52:875-884.

[35]ZHENG Guo,ZHU Wenyue,LI Xuejun,et al. Comparing diversity of canopy theridiids in major forest types in Xishuangbanna region[J]. Journal of Shenyang Normal University (Natural Science), 2013,31(4): 442-446.

[36]FLOREN A,MUELLER T,DEELEMAN-REINHOLD C,et al. Effects of forest fragmentation on canopy spider communities in SE-Asian rain forests[J]. Ecotropica, 2011,17:15-26.

[37]PURCHART L,TUF I H,HULA V,et al. Arthropod assemblages in Norway spruce monocultures during a forest cycle-a multi-taxa approach[J]. Forest Ecology & Management, 2013,306:42-51.