Downloaded from http://rsbl.royalsocietypublishing.org/ on May 29, 2015
Global change biology
rsbl.royalsocietypublishing.org
Adult activity and temperature preference drives region-wide damselfly (Zygoptera) distributions under a warming climate
Research
Jeffrey D. Corser, Erin L. White and Matthew D. Schlesinger
Cite this article: Corser JD, White EL, Schlesinger MD. 2015 Adult activity and temperature preference drives region-wide damselfly (Zygoptera) distributions under a warming climate. Biol. Lett. 11: 20150001. http://dx.doi.org/10.1098/rsbl.2015.0001
Received: 2 January 2015 Accepted: 24 March 2015
New York Natural Heritage Program, SUNY College of Environmental Science and Forestry, 625 Broadway, Albany, NY 12233, USA We analysed a recently completed statewide odonate Atlas using multivariate linear models. Within a phylogenetically explicit framework, we developed a suite of data-derived traits to assess the mechanistic distributional drivers of 59 species of damselflies in New York State (NYS). We found that length of the flight season (adult breeding activity period) mediated by thermal preference drives regional distributions at broad (105 km2) scales. Species that had longer adult flight periods, in conjunction with longer growing seasons, had significantly wider distributions. These intrinsic traits shape species’ responses to changing climates and the mechanisms behind such range shifts are fitness-based metapopulation processes that adjust phenology to the prevailing habitat and climate regime through a photoperiod filter.
Subject Areas: ecology
1. Introduction Keywords: citizen science, climate change, damselflies, distributions, New York State, phenology
Author for correspondence: Jeffrey D. Corser e-mail: [email protected]
Electronic supplementary material is available at http://dx.doi.org/10.1098/rsbl.2015.0001 or via http://rsbl.royalsocietypublishing.org.
Traits-based mechanistic approaches for understanding how climate change affects insect distributions have recently emerged, and when undertaken in a phylogenetic framework, reveal the biological mechanisms underlying landscape-level patterns. This is because the traits are known to affect individual fitness and therefore metapopulation processes such as extinction and colonization [1,2]. Yet, the fitness-based models that can test for the cause of observed range changes are very difficult to construct, in part because repeated presence –absence data are required at fine spatial resolutions and across large areas, in order to compile comprehensive up-to-date species inventories. Furthermore, intimate knowledge of species’ biological traits and how they relate to fitness, as well as robust phylogenetic frameworks, are practically non-existent for most freshwater macroinvertebrates [3]. Without clear identifiable biological outcomes to climate change effects, resource management decision-making is left largely to guesswork. Illustrating this, over the past century, the climate in New York State (NYS) has warmed up considerably [4]. Remnant boreal forests embedded within temperate zones, such as the iconic Adirondack massif, are projected to have very short climate residence times and fast velocities of temperature change [5,6]. Already this has translated into meaningful changes for resource managers because plant hardiness guidance has altered significantly since 1990, and flagship native coldwater sport fishes and some boreal birds have required interventions to stave off extirpations [7,8]. We know that longer growing seasons are having myriad effects on forested ecosystems in east-central North America including range shifts of Odonata (damsel- and dragonflies), and observations of new species arriving from the south are commonplace [9,10]. Damselflies have, in fact, demonstrated some of the fastest range margin movements, nearly tracking shifting isotherms [11,12], and a preliminary traits-based analysis [13] found that cold-adapted
& 2015 The Author(s) Published by the Royal Society. All rights reserved.
Downloaded from http://rsbl.royalsocietypublishing.org/ on May 29, 2015
2 earlier and longer emergence
faster egg/larval development
shorter generation time (voltinism)
more foraging opportunity
~40° N
longer adult flight (breeding) season
greater adult longevity
increased reproductive output
higher abundance
metapopulation growth
effects on populations and communities
range expansion
community reassembly
Figure 1. Conceptual model of adult breeding season activity as a driver of distributional changes caused by longer growing seasons for a warm-adapted, biphasic, temperate zone ectotherm. The dotted line represents a photoperiodism filter which regulates the effects of warming on fitness, especially at temperate latitudes [15]; the thick arrow shows pathway of intense natural selection in Odonata [16]. (Online version in colour.) stenothermic odonates would be the ‘losers’ under modern warmer climates. However, much uncertainty remains over the geographical scales at which certain traits like dispersal ability might predominate [3], or even whether we can predict the limits that intrinsic traits impose upon distributional shifts [14]. To investigate this problem, we developed a suite of phylogenetically explicit data-derived traits in order to test the mechanistic drivers (figure 1; electronic supplementary material, figure S1) that have shaped damselfly distributions in NYS, a large (141 000 km2) well-watered and ecologically diverse region, rising from sea level to 1630 m. As predators in both aquatic and terrestrial systems, damselflies play a strong role in structuring communities, providing tangible ecosystem services in the form of insect pest control. Based on their hypothesized ability to track isotherms, we highlight the uniquely suited aspects of this taxon as barometers [12] of rapid ongoing landscape-scale climate and habitat changes.
2. Material and methods We gained an understanding of the fine-scale distributions of all odonates ever documented from NYS during our 5 year Atlas [17]. This citizen science effort involved hundreds of volunteers following a standardized protocol designed to locate breeding populations. Nearly 4400 geo-referenced surveys were conducted at approximately 2170 unique site locations (March to November, 2005 – 2009) across the state. Data were submitted to a coordinator to ensure standardization and expert verification of vouchers.
For this analysis, we focused on the 59 known damselflies, many of whom are on their overall North American range margins here (electronic supplementary material, figure S1); see [9,17] for further details. We measured distribution as the total number of individual site records. This metric integrated a species’ range extent and frequency of occurrence across the state over the 5 year period (table 1). When logarithmically transformed, it was strongly correlated to coarser measures of distribution such as latitudinal extent and number of occupied counties (r . 0.88). We coded five explanatory variables known to be related to damselfly distributions including proxy traits for dispersal, physiology, habitat and phenology. For consistency, we used the same three broad larval habitat designations (lentic, lotic and generalist) in [18] to categorize a species’ aquatic habitat type. Adult breeding affinity was scored by reviewing our raw site records, first determining which were marked by the surveyor into each of five broad habitat types: (i) seeps (lotic), (ii) streams and creeks (lotic), (iii) rivers (lotic), (iv) ponds and lakes (lentic), and (v) bogs, fens, marshes and swamps (lentic). Lentic and lotic habitat subtypes were then collapsed and species occurring in both broad types were considered generalists. We determined the number of flight days from the species’ phenology charts in [17] modified to remove extraneous exuvial records. This adult flight season is a terrestrial activity period that varies widely among species (a few days to several months) centred around the breeding cycle. Traits correlated with dispersal ability (wing loading) were taken from the literature. For thermal preference we extracted the geo-referenced elevations [17] for each species’ records and used the maximum elevation which was strongly correlated (r ¼ 0.98) with the entire occupied elevational range. So a higher maximum elevation
Biol. Lett. 11: 20150001
effects on individual fitness
rsbl.royalsocietypublishing.org
phenological effects of warming (= longer growing season)
Downloaded from http://rsbl.royalsocietypublishing.org/ on May 29, 2015
Table 1. Five proxy measurements for damselfly species’ traits known to be related to distributions. reference
max. elevation detected 2005– 2009 (m)
[13,17]
larval morphology: lentic, lotic, generalist (lentic and lotic) adult breeding affinity: lentic, lotic, generalist
[18,19] [17]
max. site record date minus min. site record date in any year, 2005 – 2009 (no. days)
[17]
max. length female hindwing (mm)/max. total body length of female (mm)
[19]
total no. of unique site date surveys of species’ presence (2005 – 2009)
[17]
physiological tolerance thermal preference habitat preference aquatic terrestrial breeding season phenology length of flight season (2005 – 2009) dispersal ability wing loading distribution (response) no. overall site records
demonstrated population viability across a much broader range of thermal microclimates compared with a stenothermic damselfly that is only found within a narrow elevation zone [13]. Using a generalized linear model with a Gaussian error distribution, we explored the strength of these five key predictors on the distribution of 54 damselflies (we omitted species with no records) documented during our Atlas. The continuous variables were untransformed, with the exception of distribution (log[number site records]). We also regressed predictors on an independently assembled dataset of pre-Atlas county-level damselfly distribution records in NYS [20]. All analyses were conducted in R v. 3.02.
3. Results Number of flight days and maximum elevation were the only significant predictor variables in our analysis. A linear regression model with only these two factors explained 71% of the variation in the distribution (F50,2 ¼ 64.04, adj. r 2 ¼ 0.71) of damselflies in NYS (figure 2). These physiophenological proxy traits for adult breeding season length and temperature preference also explained most of the preAtlas variation in distribution (F50,2 ¼ 49.61, adj. r 2 ¼ 0.65). The inclusion of family as a phylogenetic control variable did not appreciably alter any of the r 2 values. As a further check, we mapped traits onto the phylogeny (electronic supplementary material, figure S1) and though damselflies demonstrate strong biogeographic phylogenetic structuring [9], trait states were not clustered in certain clades, but instead were well represented across the tree.
4. Discussion Damselflies possess traits such as more efficient wing designs that presumably confer greater dispersability within groups, thus shaping range sizes [21]. However, we detected no significant effects of wing loading on species’ distributions. Nor did we find a significant relationship for larval or adult habitat preference, whether or not covariates and phylogeny were accounted for. By contrast, a large body of work [18,22,23] posits that stronger dispersal tendencies arise in lentic odonates because of greater habitat instability compared with lotic habitats; thus they have larger ranges and should be able
to track climate change more closely. Unlike the above studies, we retained the generalists in our analysis and still we did not find any effect of habitat type on distribution because in this system at least, dispersal does not appear to be limiting. Yet as a region, the size of NYS is large enough that differences in dispersal ability are thought to play a strong role in freshwater macroinvertebrate species’ distributions [3], and this does appear to be the case at larger continental scales [18,21 – 23]. However, at more tractable resource management-oriented scales where most decisions are made (up to approx. 105 km2), damselflies would seem to have more than enough inherent capacity to move around. Rather, we found strong support for length of the adult flight season (figures 1 and 2) as modulated by thermal preference to be the prime drivers of damselfly distributions. This is because a species’ adult breeding season phenology results from natural selection to optimize the period of activity and reproduction in a given environment (figure 1; [24]). Flight duration during the breeding season integrates many fitness-based life-history traits that are ultimately temperature driven; however damselfly reproductive activity might conceivably be limited by predators or senescence. But fig. 11.68 in [16] clearly details instead how adult survival is controlled mostly by ‘random’ weather events such that more adult activity leads to greater lifetime reproductive success. Two studies [14,23] found only weak support for flight season as a distributional driver in odonates, but eco-physiological traits were not included in the analyses, which we argue for these ectothermic insects is essential [2]. Others have found thermally mediated [1] activity-driven results for terrestrial ectotherms [25], and it is well known that thermal tolerance is a primary driver of range size for many insects, including odonates [2,12]. But because larval activity (diapause) in Odonata is also a function of photoperiod (especially in north-temperate latitudes), photoperiodism acts as a filter for the solely temperature-driven phenological outcomes (figure 1; [15,16]). It would nevertheless be instructive to test our novel findings over varying geographical areas and against finer scale traits like CTmax, known dispersal distances and finer oviposition–habitat associations, but these detailed measurements currently exist in only a few cases. As NYS’s colder boreal ecoregions slide towards a more temperate climate regime, they become largely thermally
Biol. Lett. 11: 20150001
measurement
rsbl.royalsocietypublishing.org
species trait
3
were primarily confined to places with relatively short growing seasons. Boreal odonate community reassembly can happen on fast (decadal) time scales, meaning that even large temperate-embedded boreal ecosystems in the midlatitudes could see notable changes in their insect fauna in the coming years [26].
3.5 3.0 2.5 2.0 1.5 1.0 0.5 0
Data accessibility. See [17]. Acknowledgements. We thank all who contributed their field data to the 50
150
200
Figure 2. Semi-log scatterplot of the number of flight days (adult breeding season length) versus damselfly distribution in NYS (2005 – 2009). Each data point is a species. unsuitable for many of the cold-adapted specialists, forcing them into micro-refugia [7,8]. This pattern was evident during our surveys because the boreal (rear edge) damselflies
Atlas. We also thank Thomas ‘Nick’ Donnelly, Paul Novak and Jason Bried for their advice early on. Two anonymous reviewers gave helpful feedback.
Funding statement. Funding for the Atlas data collection and coordination was provided through NYS Wildlife grant T-2–1 in cooperation with the US Fish and Wildlife Service. Author contributions. E.L.W. coordinated the Atlas, J.D.C., M.D.S., E.L.W. collected and databased field data; M.D.S. carried out the statistical analyses, J.D.C. conceived of, and designed the study and wrote the manuscript. Conflict of interests. No competing interests.
References 1.
2.
3.
4.
5.
6.
7.
8.
Calosi P, Bilton JT, Spicer JI. 2008 Thermal tolerance, acclimatory capacity, and vulnerability to global climate change. Biol. Lett. 4, 99 –102. (doi:10.1098/ rsbl.2007.0408) Bozinovic F, Calosi P, Spicer JI. 2011 Physiological correlates of geographic range in animals. Annu. Rev. Ecol. Evol. Syst. 42, 155 –179. (doi:10.1146/ annurev-ecolsys-102710-145055) Heino J. 2011 A macroecological perspective of diversity patterns in the freshwater realm. Freshw. Biol. 56, 1703 –1722. (doi:10.1111/j.1365-2427. 2011.02610.x) Rosenzweig C, Solecki W, DeGaetano A, O’Grady M, Hassol S, Grabhorn P (eds). 2011 NYSERDA ClimAID Team, responding to climate change in New York State, the synthesis report of the Integrated Assessment for Effective Climate Change Adaptation Strategies in New York State. Albany, NY: NYSERDA. Loarie SR, Duffy PB, Hamilton H, Asner GP, Field CB, Ackerly DD. 2009 The velocity of climate change. Science 462, 1052 –1055. (doi:10.1038/ nature08649) Zhu K, Woodall CW, Clark JS. 2011 Failure to migrate: lack of tree range expansion in response to climate change. Glob. Change Biol. 18, 1042 –1052. (doi:10.1111/j.1365-2486.2011.02571.x) Warren DR, Robinson JM, Josephson DC, Sheldon DR, Kraft CE. 2012 Elevated summer temperatures delay spawning and reduce redd construction for resident brook trout (Salvelinus fontinalis). Glob. Change Biol. 18, 1804–1811. (doi:10.1111/j.13652486.2012.02670.x) Zuckerberg BA, Woods M, Porter WF. 2010 Poleward shifts in breeding bird distributions in New York State. Glob. Change Biol. 15, 1866–1883. (doi:10. 1111/j.1365-2486.2009.01878.x)
9.
10. 11.
12.
13.
14.
15.
16. 17.
18.
Corser JD, White EL, Schlesinger MD. 2014 Odonata origins, biogeography, and diversification in an Eastern North American hotspot: multiple pathways to high temperate forest insect diversity. Insect Con. Div. 7, 393 –404. (doi:10.1111.icad.12065) Craves JA. 2006 Archilestes grandis: new for Michigan. Gt Lakes Entomol. 39, 88 –90. Hickling R, Roy DB, Hill JK, Fox R, Thomas CD. 2006 The distribution of a wide range of taxa are expanding polewards. Glob. Change Biol. 12, 450 –455. (doi:10.1111/j.1365-2486.2006.01116.x) Hassall C. In press. Odonata as candidate macroecological barometers for global climate change. Freshw. Sci. 34. Rosset V, Oertli B. 2011 Freshwater biodiversity under climate warming pressure: identifying the winners and losers in temperate standing water bodies. Biol. Conserv. 144, 2311 –2319. (doi:10. 1016/j.biocon.2011.06.009) Angert AL, Crozier LG, Rissler LJ, Gilman SE, Tewksbury JJ, Chunco AJ. 2011 Do species’ traits predict recent shifts at expanding range edges? Ecol. Lett. 14, 677–689. (doi:10.1111/j.1461-0248.2011.01620.x) Saikkonen K, Taulavuori K, Hyvonen T, Gundel PE, Hamilton CE, Vanninen I, Nissinen A, Helander M. 2012 Climate change-driven species’ range shifts filtered by photoperiodism. Nat. Clim. Change 2, 239 –242. (doi:10.1038/NCLIMATE1430) Corbet PS. 1999 Dragonflies: behavior and ecology of Odonata. Ithaca, NY: Cornell University Press. White EL, Corser JD, Schlesinger MD. 2010 The New York dragonfly and damselfly survey 2005 – 2009: the distribution and status of the Odonates of New York. Albany, NY: Natural Heritage Program. Hof C, Brandle M, Brandl R. 2006 Lentic odonates have larger and more northern ranges than lotic
19.
20.
21.
22.
23.
24.
25.
26.
species. J. Biogeograph. 33, 63– 70. (doi:10.1111/j. 1365-2699.2005.01358.x) Westfall MJ, May ML. 2006 Damselflies of North America, revised edition. Gainesville, FL: Scientific Publishers. Donnelly TW. 2004 Distribution of North American Odonata. III. Calopterygidae, Lestidae, Coenagrionidae, Protoneuridae, Platysticidae. Bull. Am. Odonat. 8, 33 –99. Swaegers J, Janssens SB, Ferreira S, Watts P, Mergeay J, McPeek MA, Stoks R. 2014 Ecological and evolutionary drivers of range size in Coenagrion damselflies. J. Evol. Biol. 27, 2386 –2395. (doi:10. 1111/jeb.12481) Hof C, Brandle M, Dehling DM, Munguia M, Brandl R, Araujo MB, Rahbek C. 2012 Habitat stability affects dispersal and the ability to track climate change. Biol. Lett. 7, 565– 573. (doi:10.1098/rsbl. 2012.0023) Grewe Y, Hof C, Dehling DM, Brandl R, Brandle M. 2013 Recent range shifts of European dragonflies provide support for an inverse relationship between habitat predictability and dispersal. Glob. Ecol. Biogeograph. 22, 403–409. (doi:10.1111/ geb.12004) Chuine I. 2010 Why does phenology drive species distribution? Phil. Trans R. Soc. B 365, 3149–3160. (doi:10.1098/rstb.2010.0142) Sinervo B et al. 2010 Erosion of lizard diversity by climate change and altered thermal niches. Science 328, 894–899. (doi:10.1126/science.1184695) Flenner I, Sahlen G. 2008 Dragonfly community re-organization in boreal forest lakes: rapid species turnover driven by climate change? Insect Conserv. Div. 1, 169–179. (doi:10.1111/j.1752-4598.2008. 00020.x)
Biol. Lett. 11: 20150001
100 flight days
4
rsbl.royalsocietypublishing.org
distribution (log of site/date combinations)
Downloaded from http://rsbl.royalsocietypublishing.org/ on May 29, 2015
Global change biology
rsbl.royalsocietypublishing.org
Adult activity and temperature preference drives region-wide damselfly (Zygoptera) distributions under a warming climate
Research
Jeffrey D. Corser, Erin L. White and Matthew D. Schlesinger
Cite this article: Corser JD, White EL, Schlesinger MD. 2015 Adult activity and temperature preference drives region-wide damselfly (Zygoptera) distributions under a warming climate. Biol. Lett. 11: 20150001. http://dx.doi.org/10.1098/rsbl.2015.0001
Received: 2 January 2015 Accepted: 24 March 2015
New York Natural Heritage Program, SUNY College of Environmental Science and Forestry, 625 Broadway, Albany, NY 12233, USA We analysed a recently completed statewide odonate Atlas using multivariate linear models. Within a phylogenetically explicit framework, we developed a suite of data-derived traits to assess the mechanistic distributional drivers of 59 species of damselflies in New York State (NYS). We found that length of the flight season (adult breeding activity period) mediated by thermal preference drives regional distributions at broad (105 km2) scales. Species that had longer adult flight periods, in conjunction with longer growing seasons, had significantly wider distributions. These intrinsic traits shape species’ responses to changing climates and the mechanisms behind such range shifts are fitness-based metapopulation processes that adjust phenology to the prevailing habitat and climate regime through a photoperiod filter.
Subject Areas: ecology
1. Introduction Keywords: citizen science, climate change, damselflies, distributions, New York State, phenology
Author for correspondence: Jeffrey D. Corser e-mail: [email protected]
Electronic supplementary material is available at http://dx.doi.org/10.1098/rsbl.2015.0001 or via http://rsbl.royalsocietypublishing.org.
Traits-based mechanistic approaches for understanding how climate change affects insect distributions have recently emerged, and when undertaken in a phylogenetic framework, reveal the biological mechanisms underlying landscape-level patterns. This is because the traits are known to affect individual fitness and therefore metapopulation processes such as extinction and colonization [1,2]. Yet, the fitness-based models that can test for the cause of observed range changes are very difficult to construct, in part because repeated presence –absence data are required at fine spatial resolutions and across large areas, in order to compile comprehensive up-to-date species inventories. Furthermore, intimate knowledge of species’ biological traits and how they relate to fitness, as well as robust phylogenetic frameworks, are practically non-existent for most freshwater macroinvertebrates [3]. Without clear identifiable biological outcomes to climate change effects, resource management decision-making is left largely to guesswork. Illustrating this, over the past century, the climate in New York State (NYS) has warmed up considerably [4]. Remnant boreal forests embedded within temperate zones, such as the iconic Adirondack massif, are projected to have very short climate residence times and fast velocities of temperature change [5,6]. Already this has translated into meaningful changes for resource managers because plant hardiness guidance has altered significantly since 1990, and flagship native coldwater sport fishes and some boreal birds have required interventions to stave off extirpations [7,8]. We know that longer growing seasons are having myriad effects on forested ecosystems in east-central North America including range shifts of Odonata (damsel- and dragonflies), and observations of new species arriving from the south are commonplace [9,10]. Damselflies have, in fact, demonstrated some of the fastest range margin movements, nearly tracking shifting isotherms [11,12], and a preliminary traits-based analysis [13] found that cold-adapted
& 2015 The Author(s) Published by the Royal Society. All rights reserved.
Downloaded from http://rsbl.royalsocietypublishing.org/ on May 29, 2015
2 earlier and longer emergence
faster egg/larval development
shorter generation time (voltinism)
more foraging opportunity
~40° N
longer adult flight (breeding) season
greater adult longevity
increased reproductive output
higher abundance
metapopulation growth
effects on populations and communities
range expansion
community reassembly
Figure 1. Conceptual model of adult breeding season activity as a driver of distributional changes caused by longer growing seasons for a warm-adapted, biphasic, temperate zone ectotherm. The dotted line represents a photoperiodism filter which regulates the effects of warming on fitness, especially at temperate latitudes [15]; the thick arrow shows pathway of intense natural selection in Odonata [16]. (Online version in colour.) stenothermic odonates would be the ‘losers’ under modern warmer climates. However, much uncertainty remains over the geographical scales at which certain traits like dispersal ability might predominate [3], or even whether we can predict the limits that intrinsic traits impose upon distributional shifts [14]. To investigate this problem, we developed a suite of phylogenetically explicit data-derived traits in order to test the mechanistic drivers (figure 1; electronic supplementary material, figure S1) that have shaped damselfly distributions in NYS, a large (141 000 km2) well-watered and ecologically diverse region, rising from sea level to 1630 m. As predators in both aquatic and terrestrial systems, damselflies play a strong role in structuring communities, providing tangible ecosystem services in the form of insect pest control. Based on their hypothesized ability to track isotherms, we highlight the uniquely suited aspects of this taxon as barometers [12] of rapid ongoing landscape-scale climate and habitat changes.
2. Material and methods We gained an understanding of the fine-scale distributions of all odonates ever documented from NYS during our 5 year Atlas [17]. This citizen science effort involved hundreds of volunteers following a standardized protocol designed to locate breeding populations. Nearly 4400 geo-referenced surveys were conducted at approximately 2170 unique site locations (March to November, 2005 – 2009) across the state. Data were submitted to a coordinator to ensure standardization and expert verification of vouchers.
For this analysis, we focused on the 59 known damselflies, many of whom are on their overall North American range margins here (electronic supplementary material, figure S1); see [9,17] for further details. We measured distribution as the total number of individual site records. This metric integrated a species’ range extent and frequency of occurrence across the state over the 5 year period (table 1). When logarithmically transformed, it was strongly correlated to coarser measures of distribution such as latitudinal extent and number of occupied counties (r . 0.88). We coded five explanatory variables known to be related to damselfly distributions including proxy traits for dispersal, physiology, habitat and phenology. For consistency, we used the same three broad larval habitat designations (lentic, lotic and generalist) in [18] to categorize a species’ aquatic habitat type. Adult breeding affinity was scored by reviewing our raw site records, first determining which were marked by the surveyor into each of five broad habitat types: (i) seeps (lotic), (ii) streams and creeks (lotic), (iii) rivers (lotic), (iv) ponds and lakes (lentic), and (v) bogs, fens, marshes and swamps (lentic). Lentic and lotic habitat subtypes were then collapsed and species occurring in both broad types were considered generalists. We determined the number of flight days from the species’ phenology charts in [17] modified to remove extraneous exuvial records. This adult flight season is a terrestrial activity period that varies widely among species (a few days to several months) centred around the breeding cycle. Traits correlated with dispersal ability (wing loading) were taken from the literature. For thermal preference we extracted the geo-referenced elevations [17] for each species’ records and used the maximum elevation which was strongly correlated (r ¼ 0.98) with the entire occupied elevational range. So a higher maximum elevation
Biol. Lett. 11: 20150001
effects on individual fitness
rsbl.royalsocietypublishing.org
phenological effects of warming (= longer growing season)
Downloaded from http://rsbl.royalsocietypublishing.org/ on May 29, 2015
Table 1. Five proxy measurements for damselfly species’ traits known to be related to distributions. reference
max. elevation detected 2005– 2009 (m)
[13,17]
larval morphology: lentic, lotic, generalist (lentic and lotic) adult breeding affinity: lentic, lotic, generalist
[18,19] [17]
max. site record date minus min. site record date in any year, 2005 – 2009 (no. days)
[17]
max. length female hindwing (mm)/max. total body length of female (mm)
[19]
total no. of unique site date surveys of species’ presence (2005 – 2009)
[17]
physiological tolerance thermal preference habitat preference aquatic terrestrial breeding season phenology length of flight season (2005 – 2009) dispersal ability wing loading distribution (response) no. overall site records
demonstrated population viability across a much broader range of thermal microclimates compared with a stenothermic damselfly that is only found within a narrow elevation zone [13]. Using a generalized linear model with a Gaussian error distribution, we explored the strength of these five key predictors on the distribution of 54 damselflies (we omitted species with no records) documented during our Atlas. The continuous variables were untransformed, with the exception of distribution (log[number site records]). We also regressed predictors on an independently assembled dataset of pre-Atlas county-level damselfly distribution records in NYS [20]. All analyses were conducted in R v. 3.02.
3. Results Number of flight days and maximum elevation were the only significant predictor variables in our analysis. A linear regression model with only these two factors explained 71% of the variation in the distribution (F50,2 ¼ 64.04, adj. r 2 ¼ 0.71) of damselflies in NYS (figure 2). These physiophenological proxy traits for adult breeding season length and temperature preference also explained most of the preAtlas variation in distribution (F50,2 ¼ 49.61, adj. r 2 ¼ 0.65). The inclusion of family as a phylogenetic control variable did not appreciably alter any of the r 2 values. As a further check, we mapped traits onto the phylogeny (electronic supplementary material, figure S1) and though damselflies demonstrate strong biogeographic phylogenetic structuring [9], trait states were not clustered in certain clades, but instead were well represented across the tree.
4. Discussion Damselflies possess traits such as more efficient wing designs that presumably confer greater dispersability within groups, thus shaping range sizes [21]. However, we detected no significant effects of wing loading on species’ distributions. Nor did we find a significant relationship for larval or adult habitat preference, whether or not covariates and phylogeny were accounted for. By contrast, a large body of work [18,22,23] posits that stronger dispersal tendencies arise in lentic odonates because of greater habitat instability compared with lotic habitats; thus they have larger ranges and should be able
to track climate change more closely. Unlike the above studies, we retained the generalists in our analysis and still we did not find any effect of habitat type on distribution because in this system at least, dispersal does not appear to be limiting. Yet as a region, the size of NYS is large enough that differences in dispersal ability are thought to play a strong role in freshwater macroinvertebrate species’ distributions [3], and this does appear to be the case at larger continental scales [18,21 – 23]. However, at more tractable resource management-oriented scales where most decisions are made (up to approx. 105 km2), damselflies would seem to have more than enough inherent capacity to move around. Rather, we found strong support for length of the adult flight season (figures 1 and 2) as modulated by thermal preference to be the prime drivers of damselfly distributions. This is because a species’ adult breeding season phenology results from natural selection to optimize the period of activity and reproduction in a given environment (figure 1; [24]). Flight duration during the breeding season integrates many fitness-based life-history traits that are ultimately temperature driven; however damselfly reproductive activity might conceivably be limited by predators or senescence. But fig. 11.68 in [16] clearly details instead how adult survival is controlled mostly by ‘random’ weather events such that more adult activity leads to greater lifetime reproductive success. Two studies [14,23] found only weak support for flight season as a distributional driver in odonates, but eco-physiological traits were not included in the analyses, which we argue for these ectothermic insects is essential [2]. Others have found thermally mediated [1] activity-driven results for terrestrial ectotherms [25], and it is well known that thermal tolerance is a primary driver of range size for many insects, including odonates [2,12]. But because larval activity (diapause) in Odonata is also a function of photoperiod (especially in north-temperate latitudes), photoperiodism acts as a filter for the solely temperature-driven phenological outcomes (figure 1; [15,16]). It would nevertheless be instructive to test our novel findings over varying geographical areas and against finer scale traits like CTmax, known dispersal distances and finer oviposition–habitat associations, but these detailed measurements currently exist in only a few cases. As NYS’s colder boreal ecoregions slide towards a more temperate climate regime, they become largely thermally
Biol. Lett. 11: 20150001
measurement
rsbl.royalsocietypublishing.org
species trait
3
were primarily confined to places with relatively short growing seasons. Boreal odonate community reassembly can happen on fast (decadal) time scales, meaning that even large temperate-embedded boreal ecosystems in the midlatitudes could see notable changes in their insect fauna in the coming years [26].
3.5 3.0 2.5 2.0 1.5 1.0 0.5 0
Data accessibility. See [17]. Acknowledgements. We thank all who contributed their field data to the 50
150
200
Figure 2. Semi-log scatterplot of the number of flight days (adult breeding season length) versus damselfly distribution in NYS (2005 – 2009). Each data point is a species. unsuitable for many of the cold-adapted specialists, forcing them into micro-refugia [7,8]. This pattern was evident during our surveys because the boreal (rear edge) damselflies
Atlas. We also thank Thomas ‘Nick’ Donnelly, Paul Novak and Jason Bried for their advice early on. Two anonymous reviewers gave helpful feedback.
Funding statement. Funding for the Atlas data collection and coordination was provided through NYS Wildlife grant T-2–1 in cooperation with the US Fish and Wildlife Service. Author contributions. E.L.W. coordinated the Atlas, J.D.C., M.D.S., E.L.W. collected and databased field data; M.D.S. carried out the statistical analyses, J.D.C. conceived of, and designed the study and wrote the manuscript. Conflict of interests. No competing interests.
References 1.
2.
3.
4.
5.
6.
7.
8.
Calosi P, Bilton JT, Spicer JI. 2008 Thermal tolerance, acclimatory capacity, and vulnerability to global climate change. Biol. Lett. 4, 99 –102. (doi:10.1098/ rsbl.2007.0408) Bozinovic F, Calosi P, Spicer JI. 2011 Physiological correlates of geographic range in animals. Annu. Rev. Ecol. Evol. Syst. 42, 155 –179. (doi:10.1146/ annurev-ecolsys-102710-145055) Heino J. 2011 A macroecological perspective of diversity patterns in the freshwater realm. Freshw. Biol. 56, 1703 –1722. (doi:10.1111/j.1365-2427. 2011.02610.x) Rosenzweig C, Solecki W, DeGaetano A, O’Grady M, Hassol S, Grabhorn P (eds). 2011 NYSERDA ClimAID Team, responding to climate change in New York State, the synthesis report of the Integrated Assessment for Effective Climate Change Adaptation Strategies in New York State. Albany, NY: NYSERDA. Loarie SR, Duffy PB, Hamilton H, Asner GP, Field CB, Ackerly DD. 2009 The velocity of climate change. Science 462, 1052 –1055. (doi:10.1038/ nature08649) Zhu K, Woodall CW, Clark JS. 2011 Failure to migrate: lack of tree range expansion in response to climate change. Glob. Change Biol. 18, 1042 –1052. (doi:10.1111/j.1365-2486.2011.02571.x) Warren DR, Robinson JM, Josephson DC, Sheldon DR, Kraft CE. 2012 Elevated summer temperatures delay spawning and reduce redd construction for resident brook trout (Salvelinus fontinalis). Glob. Change Biol. 18, 1804–1811. (doi:10.1111/j.13652486.2012.02670.x) Zuckerberg BA, Woods M, Porter WF. 2010 Poleward shifts in breeding bird distributions in New York State. Glob. Change Biol. 15, 1866–1883. (doi:10. 1111/j.1365-2486.2009.01878.x)
9.
10. 11.
12.
13.
14.
15.
16. 17.
18.
Corser JD, White EL, Schlesinger MD. 2014 Odonata origins, biogeography, and diversification in an Eastern North American hotspot: multiple pathways to high temperate forest insect diversity. Insect Con. Div. 7, 393 –404. (doi:10.1111.icad.12065) Craves JA. 2006 Archilestes grandis: new for Michigan. Gt Lakes Entomol. 39, 88 –90. Hickling R, Roy DB, Hill JK, Fox R, Thomas CD. 2006 The distribution of a wide range of taxa are expanding polewards. Glob. Change Biol. 12, 450 –455. (doi:10.1111/j.1365-2486.2006.01116.x) Hassall C. In press. Odonata as candidate macroecological barometers for global climate change. Freshw. Sci. 34. Rosset V, Oertli B. 2011 Freshwater biodiversity under climate warming pressure: identifying the winners and losers in temperate standing water bodies. Biol. Conserv. 144, 2311 –2319. (doi:10. 1016/j.biocon.2011.06.009) Angert AL, Crozier LG, Rissler LJ, Gilman SE, Tewksbury JJ, Chunco AJ. 2011 Do species’ traits predict recent shifts at expanding range edges? Ecol. Lett. 14, 677–689. (doi:10.1111/j.1461-0248.2011.01620.x) Saikkonen K, Taulavuori K, Hyvonen T, Gundel PE, Hamilton CE, Vanninen I, Nissinen A, Helander M. 2012 Climate change-driven species’ range shifts filtered by photoperiodism. Nat. Clim. Change 2, 239 –242. (doi:10.1038/NCLIMATE1430) Corbet PS. 1999 Dragonflies: behavior and ecology of Odonata. Ithaca, NY: Cornell University Press. White EL, Corser JD, Schlesinger MD. 2010 The New York dragonfly and damselfly survey 2005 – 2009: the distribution and status of the Odonates of New York. Albany, NY: Natural Heritage Program. Hof C, Brandle M, Brandl R. 2006 Lentic odonates have larger and more northern ranges than lotic
19.
20.
21.
22.
23.
24.
25.
26.
species. J. Biogeograph. 33, 63– 70. (doi:10.1111/j. 1365-2699.2005.01358.x) Westfall MJ, May ML. 2006 Damselflies of North America, revised edition. Gainesville, FL: Scientific Publishers. Donnelly TW. 2004 Distribution of North American Odonata. III. Calopterygidae, Lestidae, Coenagrionidae, Protoneuridae, Platysticidae. Bull. Am. Odonat. 8, 33 –99. Swaegers J, Janssens SB, Ferreira S, Watts P, Mergeay J, McPeek MA, Stoks R. 2014 Ecological and evolutionary drivers of range size in Coenagrion damselflies. J. Evol. Biol. 27, 2386 –2395. (doi:10. 1111/jeb.12481) Hof C, Brandle M, Dehling DM, Munguia M, Brandl R, Araujo MB, Rahbek C. 2012 Habitat stability affects dispersal and the ability to track climate change. Biol. Lett. 7, 565– 573. (doi:10.1098/rsbl. 2012.0023) Grewe Y, Hof C, Dehling DM, Brandl R, Brandle M. 2013 Recent range shifts of European dragonflies provide support for an inverse relationship between habitat predictability and dispersal. Glob. Ecol. Biogeograph. 22, 403–409. (doi:10.1111/ geb.12004) Chuine I. 2010 Why does phenology drive species distribution? Phil. Trans R. Soc. B 365, 3149–3160. (doi:10.1098/rstb.2010.0142) Sinervo B et al. 2010 Erosion of lizard diversity by climate change and altered thermal niches. Science 328, 894–899. (doi:10.1126/science.1184695) Flenner I, Sahlen G. 2008 Dragonfly community re-organization in boreal forest lakes: rapid species turnover driven by climate change? Insect Conserv. Div. 1, 169–179. (doi:10.1111/j.1752-4598.2008. 00020.x)
Biol. Lett. 11: 20150001
100 flight days
4
rsbl.royalsocietypublishing.org
distribution (log of site/date combinations)
Downloaded from http://rsbl.royalsocietypublishing.org/ on May 29, 2015
