Sunteți pe pagina 1din 18

Freshwater Biology (2007) 52, 796–813

doi:10.1111/j.1365-2427.2007.01733.x

Functional diversity of crustacean zooplankton communities: towards a trait-based classification

ALLAIN J. BARNETT, KERRI FINLAY AND BEATRIX E. BEISNER Department of Biological Sciences, University of Quebec at Montreal, Montreal, QC, Canada

SUMMARY

1. While studies of phytoplankton and terrestrial plant communities have increasingly

emphasised the use of functional traits in ecological research, few have yet to apply this

approach to zooplankton communities.

2. This study reviews laboratory and observational studies on zooplankton feeding and life

history and provides a series of functional trait tables for the North American freshwater zooplankton. Qualitative and quantitative trait tables highlight areas where data were

more scarce and point to which types of studies could fill in gaps in our knowledge of zooplankton niches.

3. Data were most complete for the Cladocera across most traits, while feeding information

for cyclopoids was most sparse. Qualitative data that distinguished congeneric species were lacking for most groups.

4. A regional community dendrogram for common north-eastern North American

zooplankton species was generated and shows that taxonomic differences between species

do not capture fully functional differences based on the traits of body length, habitat, trophic group and feeding type.

5. The data collected here, combined with readily measurable species attributes, can be

used to generate a multivariate measure of the functional niche of each species found in a

community. Armed with this information, functional relationships that are useful for ecological studies of lake ecosystems can be more easily conducted.

Keywords: biodiversity, classification, functional diversity, niche, traits

simply the number of taxonomic species (Walker, 1991; Hooper & Vitousek, 1997; Tilman et al. , 1997; Walker, Kinzig & Langridge, 1999; Symstad, Siemann & Haarstad, 2000). Thus, there have been recent calls for the use of functional diversity measures – descrip- tors of communities based on a new type of classifi- cation scheme that is applicable to many ecological questions (Petchey & Gaston, 2002; Norberg, 2004; Mason et al., 2005; Mouillot et al., 2005) and inspired by previous approaches (e.g. Grime, 1977; Reynolds, 1980) including the niche concept. Functional diversity is a biodiversity measure based on functional traits of the species present in a community. Functional traits are those that define species in terms of their ecological roles – how they interact with the environment and with other species (Diaz & Cabido, 2001). In phytoplankton, for example,

Introduction

The human proclivity towards the classification of nature has led to the definition and cataloguing of species in communities based traditionally on mor- phological and more recently, on genetic descriptors. While this taxonomic classification is essential, it may not be the most appropriate scheme for many ecolog- ical applications. For example, many studies on the relationship between biodiversity and ecosystem function have concluded that it is the ecological roles of the species present that are important, and not

Correspondence: Beatrix E. Beisner, Department of Biological Sciences, University of Quebec at Montreal, CP 8888, Succ. Centre-Ville, Montreal, QC, Canada H3C 3P8. E-mail: beisner.beatrix@uqam.ca

796

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd

these traits usually include body size, tolerance and sensitivity to environmental conditions, motility,

shape, N-fixation ability and Si requirements (Rey- nolds et al., 2002; Weithoff, 2003). In terrestrial plant communities, researchers have included more com- plex traits like rates of growth, nutrient requirements and water uptake (Walker & Langridge, 2002). Similarly, in stream environments, a functional approach to invertebrate communities have been invoked for examining diversity patterns and for biomonitoring purposes (e.g. Statzner et al., 2001; Bady et al., 2005). There are several methods for calculating functional diversity of a community. Recent methods include Functional Attribute Diversity, as used in a study of Australian rangelands by Walker et al. (1999), and Functional Diversity (FD) proposed more recently by Petchey & Gaston (2002). Trait variance, measured as the width of a trait distribution, has been proposed by Norberg (2004). Beyond the simple measurement of diversity, Mason et al. (2005) proposed also estimating functional richness, functional evenness and func- tional divergence, to enable descriptions of niche use and competitive interactions in communities. In order to take a functional approach and to use these new measures, however, we must have descriptors of the functional groups present in a community. The effects of environmental changes on ecosystem properties are thought to be largely determined by the range of biological processes carried out by the suite

of species present in the community (Reynolds et al. ,

2002). Studies that focus on whether biodiversity affects the functioning of an ecosystem (e.g. produc-

tivity, total biomass) have yielded results that suggest

a link between the functional roles of species and

community responses to environmental changes. For example, Tilman et al. (1997) found that measures of functional diversity were more strongly related to ecosystem response variables (e.g. productivity, soil NO 3 , soil NH 4 ) in a plant community than was traditional species diversity. In another study, Walker et al. (1999) observed that minor, functionally redund- ant species in ungrazed grasslands increased in abundance in heavily grazed communities, providing support for the insurance hypothesis (McNaughton, 1977; Yachi & Loreau, 1999), whereby seemingly redundant species contribute to ecosystem resilience when environmental conditions change. Functional traits of species describe how they respond to the

Functional diversity of zooplankton communities 797

environment, and if we know what functional traits

are present in a community, then we should be better

able to predict how communities will shift as envi- ronmental conditions change and thereby understand

the role of biodiversity more fully (Norberg, 2004).

A functional classification of species should greatly enhance our ability to predict how the pulsed and gradual changes of environmental conditions will affect communities. For terrestrial plant species, a

classification strategy has recently been proposed to predict changes in vegetation in response to environ- mental change (Lavorel, McIntyre & Landsberg, 1997).

A similar strategy has also been proposed for

phytoplankton to be applied to ecological questions such as the intermediate disturbance hypothesis, insurance hypothesis and alternative stable state shifts

in lakes (Reynolds et al., 2002; Weithoff, 2003). For example, the intermediate disturbance hypothesis

predicts the highest diversity of species at intermedi-

ate magnitude or frequency of disturbance by per-

mitting species with different strategies and competitive advantages to coexist (Connell, 1978).

Weithoff (2003) asserts that this hypothesis relies on

the assumption that species are functionally dissimilar

in response to disturbance. For a greater understand-

ing of the catastrophic shift with eutrophication from

a vegetation dominated clear state to an alga domin- ated turbid one, a functional approach can be used to determine the gradual changes in attributes of the algal community in response to nutrient addition, leading up to a shift to an alternative state (Weithoff,

2003). Finally, the use of functional diversity for testing the ‘insurance hypothesis’ provides insight into the potential ‘roles’ of seemingly minor or redundant species in lake communities (Weithoff, 2003) as well as in terrestrial ones (Walker et al., 1999). Therefore, the use of functional diversity enhances our understanding of these important ecological phenomena. In aquatic environments, zooplankton play an important role in structuring phytoplankton commu- nities (McCauley & Briand, 1979; Sterner, 1989), and in mediating energy flow to higher trophic levels in pelagic habitats (Gliwicz & Pijanowska, 1989). Until now, no thorough examination of the potential for functional classification of zooplankton has been attempted. We explore the possibility of determining functional grouping strategies for zooplankton using

a review of the literature to determine for which

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

798 A.J. Barnett et al.

groups and traits knowledge exists and for which information is still lacking. We develop here, a functional classification tree for common North American zooplankton and show that taxonomic relationships are not always sufficient surrogates for questions of function. The data reviewed and collec- ted here will be useful both to expand our under- standing of traits of zooplankton, and to determine where future studies can focus their efforts in order to achieve a more robust classification method of func- tional diversity to help predict ecosystem function in lakes.

Methods

We conducted an extensive search of the primary literature for freshwater zooplankton functional traits. Emphasis was on studies that focused on traits that related to body size, feeding, growth and predator avoidance. These traits were chosen because they describe both an organism’s response to environmental conditions (e.g. body size, life history, predator avoidance, food availability), and their potential effects on ecosystem processes (e.g. feeding rate). The investigation focused on traits for zooplankton species found in north-eastern North America, where our research on plankton communities is based. However traits for other common North American species were included in the final analysis. Searches were initiated using online databases: Biological Sciences (1960–2006) and ISI’s Web of Science (1981– 2006) using keywords ‘freshwater’, ‘zooplankton’, ‘body size’, ‘feeding’, ‘growth’, ‘development’ or simply the names of each species. The reference sections of papers found in this manner were also used to search for further data in older publications. The average adult body length of each species included in the trait table were obtained wherever possible from published studies. When only ranges of body sizes were found, we present the mid-point of this range in the results tables. For values that we were unable to obtain from published studies, we included our own mean values from zooplankton communities in 18 lakes in the Eastern Townships of Quebec, Canada. Standard length measurements of zooplank- ton (McCauley, 1984) were taken on a minimum of 10 individuals from each lake in July 2004 using an Olympus dissection microscope and Evolution Qei Monochrome camera using ImagePro software under

20 · or 32 · magnification. The overall mean across all lakes for each species was calculated. Only growth rates, feeding rates and feeding rate regressions with body length that were measured in published laboratory studies at temperatures between 15 and 25 C, were included in results tables as they reflect mean summer lake temperatures in this region of North America. For copepods, we included only rates for adult females. Information found in the text of studies that provided qualitative information, such as habitat or food preferences, or presence/absence of

a defensive strategy was also included. All data were

compiled into trait tables and separated into quanti-

tative and qualitative data tables.

Community dendrogram

Qualitative and quantitative traits that were found for

a sufficient number of species (length, habitat, trophic group and feeding type) were used to generate a

functional dendrogram for the regional species pool of north-eastern North American lake zooplankton. In addition to the full functional dendrogram based on the four characteristics, a functional dendrogram using only body length was generated for comparison of what information can be gleaned using a very limited estimate of species functional traits. Qualitative measures were entered as ranked categories (i.e. from herbivore to carnivore, and from more passive forms of filtering type through to raptorial feeding). The dendrogram of species for which sufficient traits existed was generated using hierarchical clustering analysis from standardised Euclidean distances as in the method of FD calculation proposed by Petchey & Gaston (2002). These dendrograms were generated using average linkage clustering and unweighted pair grouping (Krebs, 1999) using the statistical computer program PRIMERPRIMER (Clarke & Gorley, 2001). FD can be considered as a quantitative measure of the richness of functional traits, or the degree of complementarity within a community (Petchey & Gaston, 2002). FD for

a particular lake community of interest (containing a

subset of species presented) would be the sum of branch lengths in the dendrogram of the regional species pool calculated herein. Dendrograms were compared with a simplified taxonomic tree for the regional species pool that was constructed according to major family and class designations for each species according to Thorp & Covich (2001).

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

Results

The life history information presented was collected from 126 sources spanning Frank (1952) to Frost et al. (2006). The majority of references are from the primary peer-reviewed literature, while some (particularly qualitative) data were taken from the Great Lakes Copepod key of the USGS Great Lakes Science Center (http://www.glsc.usgs.gov/greatlakescopepods), and other taxonomic keys (e.g. Pennak, 1989; Hebert, 1995; Witty, 2004). Only traits for which data on more than 25 % of all species were found were included. A complete list of the primary literature used to generate trait tables is given in Supplementary material .

Quantitative traits defined

Quantitative traits are listed in Tables 1 and 2. Two measures of feeding were included in the trait table for both copepods and cladocerans: food size range ( lm) and maximum clearance rate ( lL ind. ) 1 h ) 1 ). Mesh size (l m) and regressions of clearance rate with body size were sufficiently abundant in the literature on cladocerans to warrant inclusion. Food size range represents the smallest and largest particles consumed by each species calculated over all cited studies. For non-spherical particles, sizes were standardised as equivalent spherical diameter ( lm ESD) using the equation (length · width) 0.5 . The studies used to calculate the range of food size included only those that measured a feeding rate explicitly, and excluded those that reported food size in relation to some other measurement, such as generation time. Whenever possible, values were from studies that assessed preferred food sizes across several species of zooplankton – the best way to collect such comparative data. Further, for Cladocera, the mesh size of the feeding appendages provided an alternative measure of the size of the smallest particle an organism is capable of capturing and this was also recorded in Table 1. Clearance rate is the volume of water swept clear per individual zooplankter per unit time. This value often remains high and constant at lower food concentrations and then declines with increasing food concentrations (Sterner, 1989; Mauchline, 1998). When multiple clearance rate values were reported in a single study, we included only the rate at the lowest food concentration in order to estimate maximum

Functional diversity of zooplankton communities 799

values better. Maximum values from each study are reported in Table 1. Because cladocerans do not undergo metamorphosis, it is often unclear for what stage a clearance rate is being measured. Moreover, it has been demonstrated that clearance rate in cladoc- erans is affected by body size. For this reason, we also included clearance rate versus body length regression relationships in Table 1. For copepods, clearance rates for adult females were used. The measures of growth included in the trait table are generation time for both cladocerans and cope- pods, and rate of population increase (r) for cladoc- erans. Generation time is defined as the number of days from hatching until moult into the adult for copepods, and from hatching until production of the first clutch in cladocerans. Generation time can vary with temperature and food concentration, so we preferentially selected maximum generation times determined at 20 C, exceptions are noted in the Tables. The rate of population increase is typically measured using either life tables or by population growth studies, and results from both types were included in the cladoceran trait table as separate columns. Stoichiometric ratios of C : P and C : N as meas- ured for individual species of zooplankton were also included in the trait table, as these values will affect nutrient excretion rate and in general capture proces- ses of material flow and cycling through the zooplank- ton.

Quantitative data results

Data for most species were obtained for mean body length, food size range and clearance rate, but were lacking for calanoid generation time and for stoi- chiometric ratios of many species. Food size range tended to vary with major taxonomic grouping since, for example, raptorially feeding cylopoids generally fed on much larger food items. Cladocer- ans, which generally filter food, fed on the smallest food particles. A large range in clearance rate was observed both between and within species. The lowest observed clearance rate was for the small herbivorous cladoceran Bosmina longirostris , with a value of 13 l L ind. ) 1 day ) 1 , while the highest was observed for the carnivorous cladoceran Polyphemus pediculus, at 20 000 lL ind. ) 1 day ) 1 . Between studies on the same species, Cyclops

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

Table 1 Quantitative functional traits of cladoceran zooplankton including sources in superscript numbers (see Supplementary material). Max r is the rate of population increase. Parameters y and x in the regression equations represent clearance rate and body size (mm), respectively. Body lengths marked with an asterisk (*) are mid-points of the range observed and those without a reference are personal observations of mean size.

6.9 [31,55 )) 57]

10.0 [42 )) 44]

C:N [30]

6.1 [37,42]

6.1 [37]

97.5 [31,55 )) 57]

140.4 [42 )) 44]

82.6 [37,42]

156.0 [37]

C:P [30]

0.28 [34] , 0.07 [35]

growth ) 1 ) expts

population

0.23 [41]

0.28 [36]a

0.32 [36]a

0.27 [34]

Max r

(day

0.22 [39]b , 0.23 [38] ,

0.44 [46] , 0.58 [47]b

0.14 [38] , 0.2 [50] , 0.29 [51] , 0.32 [52] ,

0.24 [38] , 0.29 [39] 0.31 [40]

0.36 [58] , 0.38 [59]

0.35 [53,54] , 0.37 [39]b

0.27 [48]b

Max r life table (day ) 1 )

0.33 [39]

0.27 [45]b

0.34 [33]

generation

7 [1] , 12 [50]

5 [1] , 9 [32]

7 [1] , 8 [49]

Fastest

(days)

18 [45]b

17 [32]

10 [33]§

7.2 [1]

time

6 [1]

6 [1]

0.32–1.0 [4] , 0.4–0.7 [12] ,

y ¼ 0.085 · 1.74[20] 0.23–0.45 [4]

Mesh size

distances)

0.24–0.4 [4]

0.68–1.6 [4]

0.24–0.4 [4]

0.2–0.6 [12]

y ¼ 0.487 · 1.83[15] 0.4–0.7 [12]

0.42 0.5–0.9 [6] ,

1.1–1.4 [6]

[24]

[24]

(setulae

0.38

lm)

(

y ¼ 0.26· 1.58[20]

y ¼ 0.41· 3.63[19]

ind. ) 1 h ) 1 )

length (CR in

Regression

versus

mL

CR

13 [8]a , 40 [9]b , 125 [10] , 146 [11]b ,

190 [14] , 200 [15]a , 230 [16] , 300 [17]

1.1–20 [4] 258 [22]a , 1375 [23] ,

150 [12] , 170 [13] ,

188 [18] , 321 [9]

Max clearance rate ( lL ind. ) 1 h ) 1 )

1650 350 [24] , 688 [6] ,

13 [12] , 131 [2]

[24] **

106 [10,11]

820 [25]

196 [6]

56 [2]

0.4–2 [4]

0.4–7 [4]

1.4–5 [4]

range

Food

m)

size

(l

[1] *

[1] *

0.8 [1] *

0.6 [3] *

1.7 [3] *

length

0.29 0.88 [7]

0.85 [7]

Mean

2 [21] *

0.44 0.4 [7]

(mm)

body

0.8 [7]

0.8 [7]

1.67

0.46

Ceriodaphnia lacustris

Eubosmina corregoni

Chydorus brevilabris

Bosmina longirostris

Chydorus sphaericus

Bosmina longispina

Ceriodaphnia dubia

Daphnia ambigua

Daphnia catawba

Alona rectangula

Daphnia galeata

mendotae Sars

Ceriodaphnia

Alona affinis

Scourfield

longispina

reticulata

Richard

Leydig

Leydig

Muller¨

Muller¨

Muller¨

Daphnia

O.F.M.

Species

Coker

Jurine

Baird

Frey

Sars

800 A.J. Barnett et al.

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

92.7 [55,57,65,66] 5.9 [55,57,65,66]

C:N [30]

6.4 [31,37]

7.1 [31,78]

5.7 [55]

7.3 [74]

123.5 [31,78]

115.5 [31,37]

106.0 [74]

154.0 [55]

growth ) 1 ) expts C:P [30]

population

[41]

[41]

0.046 [77] ,

0.25 0.155 [77]

0.22 0.20 [34] ,

0.29 [60]

0.25 [73]

Max r

(day

0.4 [70]b , 0.50 [71] ,

0.25 [38] , 0.40 [62] , 0.44 [63] , 0.58 [64] 0.29 [69] , 0.35 [38] ,

0.33 [50,75]

Max r life table (day ) 1 )

0.51 [72]

0.32 [79]

11 [1] , 16 [32]

generation

16 6.5 [1] , 7 [67] ,

6 [60] , 8 [1] ,

6 [1] , 9 [79]

[61]

Fastest

(days)

[68]

time

6 [76]

8 5 [50]

with D. mendotae in North America).

(gnathobasal setae) [4] ,

(gnathobasal setae) [4] ,

0.4–0.7 [12] , 0.47 [24]

Superscript letters denote values at temperatures other than 20 C. a ¼ 15–19 C, b ¼ 21–25 C.

0.16–0.24 [4] ,

2350 [24] , 3560 [26] y ¼ 0.28· 2.2[27] 0.24–0.64 [4] ,

y 2080 [24] , 2483 [23] y ¼ 0.39· 3.46[19] 0.45–1.4 [4] ,

Mesh size

distances)

1.8–3.9 [6]

0.2–0.3 [6]

¼ 0.23· 2.74[15] 0.4–0.7 [12]

[24]

[24]

(setulae

0.9–2.3

0.46 0.40 [24]

2.48

y y ¼ 0.38· 0.36[20] 1.2–2

lm)

(

¼ 0.26· 1.27[20]

¼ 0.41x 3.37[19]

y ¼ 0.42· 1.61[20] ,

¼ 0.31· 3.02[22] ,

ind. ) 1 h ) 1 )

length (CR in

hybridises complex)

Regression

versus

mL

CR

which mendotae

y
y

267 [22] a , 867 [9] , 1067 [18] , 1190 [12] , 1200 [15]

2920 [24] , 3800 [12] , 9929 [2]

0.25–5.00 [4] 213 [18] , 237 [9] , 271 [6] , 389 [10] ,

326 1920 [28] [24] , 570 [29] ,

390 [12] 406 [11]

Max clearance rate ( lL ind. ) 1 h ) 1 )

521 4979 [9] , [18] 2271 [6] ,

D. galeata

>20000 [30]

(part of the species

dentifera (European

[20] * 4.0–25 [4]

1.5–30 [4]

[21] * 0.6–40 [4]

4.5–25 [4]

range

Food

lm)

size

al. (2005)

(

Daphnia galeata

data. (2000). ‡ † As Based cited on in eld Gillooly

Lynch et (1980)

length

[7]

[7]

1.75 [7]

Mean

(mm)

body

2.55 0.83*

1.75 1.4 [7]

1.04

0.85

†† Data are are for for Daphnia

3.5

Table 1 ( Continued)

cited in in Sarma

pediculus Linnee

Daphnia pulicaria

Daphnia magna

Sida crystallina

Leydig pulex

Diaphanosoma

Daphnia rosea

brachyurum

Holopedium

Polyphemus

Zaddach

gibberum

§ As As cited

´Lieven

Muller¨

Daphnia

Forbes

Straus

Species

**Data

Sars

Functional diversity of zooplankton communities 801

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

5.2 [31,98]

5.2 [31]

5.7 [31]

5.8 [31]

C : N

92.4 [31,98]

116.6 [31]

107.6 [31]

131.3 [31]

C : P

13.4 [81] , 13.9 [82] , 14.8 [83]a , 16.5 [84] , 18 [85] ,19.8 [86]

17.2 [101] 16.8 [103]a , 41.4 [104] 9.7 [101]b , 23.5 [81] , 27.7 [86]

23 [81] , 25.1 [82]b , 26.3 [94] ,

Fastest generation

27 [86] , 31 [95]

time (days)

12 [86]

Table 2 Quantitative functional traits of copepod species including sources in superscript numbers (see Supplementary material).

487 [106] , 571 [102] , 790 [11]b , 805 [10] 1356 [102] , 1666 [108] 667 [102] , 966 [105]

28.5 [92]a , 59 [93] , 1200 [91]a

213 [97]a , 7500 [96]a

Max CR ( lL ind. ) 1 h ) 1 )

Superscript letters denote values at temperatures other than 20 C. a ¼ 15–19 C, b ¼ 21–25 C.

2708 [100]

380 [102]

6.5–80 [10,11,102,105,106]

<900 lm rotifers [99] 6.5–80 [102]

30–80 [102,105,109]

6–31.7 [89 )) 91]

Food size range (l m)

3–80 [102,107]

114–562 [80]

0.25–5.2 [87]

2.5–30 [107]

15–100 [96]

5–50 [110]

length (mm)

Mean body

0.89 [102]

0.55 [102]

1.44 [102]

1.14 [102]

1.63 [88]

0.89 [88]

1.07

0.84

0.84

1.14

1.06

1.26

0.72

0.80

0.88

0.93

1.53

Diacyclops bicuspifatus thomas Forbes Eucyclops speratus Lilljeborg Mesocyclops edax Forbes Tropocyclops prasinus Fischer Thermocyclops crassus Fischer Calanoida Diaptomus minutus Lilljeborg Diaptomus oregonensis Lilljeborg Epischura lacustris Forbes Leptodiaptomus minutus Lilljeborg Leptodiaptomus sicilis Forbes Onychodiaptomus birgei Marsh Skistodiaptomus oregonensis Lilljeborg Skistodiaptomus reighardi Marsh

Acanthocyclops vernalis Fischer Cyclops scutifer Sars Cyclops vicinus Ulianine

robustus Sars

Acanthocyclops

Cyclopoida

Species

802 A.J. Barnett et al.

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

vicinus was observed to feed at 28–63 lL ind. ) 1 day ) 1 on algae (Adrian, 1991; Santer, 1993) and 1200 lL ind. ) 1 day ) 1 on rotifers (Rabette, Thouvenot & Lair, 1998). Generation time tended to be shorter for cladocer- ans than cyclopoids, but again observed values varied considerably both between and within species. The most extreme example was for Thermocylops crassus in which generation time ranged from 9.7 days (Wyng- aard et al., 2005) to 27.7 days (Maier, 1994). Stoichiometric ratios (C : P, C : N) for freshwater zooplankton have recently been summarised in two major studies (Elser et al., 2000; Frost et al., 2006) and these values were included here. Unfortunately, only values for common cladoceran species and four copepod species appear to have been determined so far.

Qualitative traits defined

Qualitative traits found that were sufficiently com- mon to include in Tables 3 and 4 were habitat, trophic group, optimal productivity conditions, feeding type and selectivity, for both copepods and cladocerans, and preferred food size category, predatory escape response and helmet forming capacity, for cladocer- ans only. The habitat in which each organism was most likely to be found was categorised as either littoral or pelagic. Trophic group was divided into the traditional herbivore, omnivore and carnivore categ- ories with the addition of herbivore–omnivore and carnivore–omnivore. This addition was used to dis- tinguish between copepods such as Mesocyclops edax , that is more carnivorous, and Tropocyclops prasinus that is more herbivorous. In this study, feeding on non-photosynthetic protists was considered herbivo- ry, in order to distinguish between those species that feed on protists (e.g. Daphnia spp.), and those that feed on larger organisms like rotifers and crustaceans (e.g. Polyphemus spp. and Mesocyclops spp.). Optimal pro- ductivity conditions characterised the lake trophic status at which optimal competitive ability is achieved by each species. For example, chydorids are more successful in very productive lakes while daphniids are generally more successful in low productivity lakes with high quality food (Tables 3 and 4). Food size was used to describe broadly the relative food size category preferred by each species. Predatory escape response characterised how a species evades

Functional diversity of zooplankton communities 803

an attack from a predator. The ability to produce a helmet in Cladocera was also included as an anti- predation trait. Feeding type is the manner in which species or groups obtain their food. Cladocerans have been divided into four classes based on DeMott & Kerfoot (1982); (1) Daphnia-type (D-type), where filtering is from a stationary position with filtering apparatus on the third and fourth legs, (2) Sida-type (S-type), similar to D-type with the exception that the filtering appar- atus is located on the first five legs, (3) Bosmina -type (B-type), characterised by a horizontal active swim- ming and less developed filtering apparatus on thoracic appendages and (4) Chydorus -type (C-Filtra- tion) where feeding is predominantly by scraping algal particles from periphyton. Amongst the cope- pods, raptorial predators like cyclopoids actively capture and kill prey, while stationary suspension feeders are more passive with less frequent swim- ming. Finally Epischura lacustris is differentiated from both raptorial and stationary suspension feeding as it swims continuously while creating feeding currents, thus being characterised as a ‘current cruiser’. The last qualitative trait in Tables 3 and 4 is selectivity, a relative estimation of how selective an organism is towards its food. Raptorial feeders are generally considered to be more selective than filter feeders, but there are also relative degrees of selec- tivity exhibited within the filter feeding Cladocera species. The degree of selectivity can be defined based on how active a species is in seeking out food of a particular quality, through selective filtering or particle rejection.

Qualitative data results

The qualitative trait matrix contains a number of trait variables available only for some taxonomic groups (Tables 3 and 4). For example, the ability to form helmets obviously does not apply to copepods. Based on our biological knowledge of many of these species, qualitative traits were often assumed to be the same within a genus, or sometimes within a family or order. The majority of trait information in Tables 3 and 4 is considered to apply to the entire genus, although exceptions mentioned specifically in the literature are cited. The preferred food of most cladocerans is algae, and thus there is no trophic group distinction given between cladoceran species as all are thought to

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

Moderate [15,16]

Moderate [15,16]

Moderate [15,16]

Selectivity

Medium [5] Yes [21,117] D-Filtration [41] Low [15,16]

Low [15,16]

D-Filtration [41] Low [15,16] D-Filtration [41] Low [15,16] D-Filtration [41] Low [15,16] D-Filtration [41] Low [15,16] D-Filtration [41] Low [15,16] S-Filtration [41] S-Filtration [41] S-Filtration [41] Raptorial [30]

Low [14,15]

with a ‘ ’. numbers (see Supplementary material). Traits without a reference are personal

D-Filtration [41]

D-Filtration [41]

D-Filtration [41]

Yes [21,117] D-Filtration [41]

B-Filtration [41]

B-Filtration [41]

B-Filtration [41]

D-Filtration [41]

Feeding type

C-Filtration

C-Filtration

C-Filtration

No [21,117]

Helmet

No [117]

No [21]

Medium [5] No [21]

No [21]

No

No

No

No

No

No

No

No

No

No

No

No

No

No

Food size

Large [5]

Large [5]

Large [5]

Large [5]

Large [5]

Large [5]

Small [5]

Small [5]

All [5]

All [5]

Low [29,100,116]

Low [29,114,116]

productivity

Low [29,116]

Low [29,116]

Low [29,116]

conditions

Low [29,116]

Low [29,116]

Herbivore [113] High [36]

High [36]

Herbivore [113] High [114]

High [114]

Optimal

Herbivore [113]

Herbivore [113]

Herbivore [113]

Herbivore [113]

Herbivore [113]

Herbivore [113]

Herbivore [113]

Herbivore [113]

Herbivore [113]

Herbivore [113]

Pelagic [1,113] Herbivore [15]

Herbivore [15]

Herbivore [15]

Herbivore [112]

Herbivore [114]

Carnivore [30]

Herbivore [2]

Herbivore [2]

Herbivore [2]

indicated in superscript

Trophic

group

Littoral [1,115]

Pelagic [1,113]

Pelagic [1,113]

Pelagic [1,113]

Pelagic [1,113]

Pelagic [1,113]

Littoral [113]

Pelagic [115]

Pelagic [113]

Pelagic [113]

Pelagic [115]

Reduced swimming [111,112] Pelagic [113] Reduced swimming [111,112] Pelagic [113]

Pelagic [115]

[1,2]

Littoral [1,2]

a group is sources

Pelagic [30]

[1]

Pelagic [3]

Ponds [21]

Habitat

Littoral¨

Littoral¨

species of including

[111,112]

Predatory escape response

Rapid swimming [112] Rapid swimming [112] Rapid swimming [112] Rapid swimming [112] Pausing and jumping [112]

[112]

Rapid swimming [112] Rapid swimming [111] Rapid swimming [112]

Rapid swimming [112] Rapid swimming [112] Rapid swimming [112]

generalisation

Reduced swimming¨

Rapid swimming¨

traits of by cladoceran

functional obtained

brachyurum Lieven ´Diaphanosoma

Holopedium gibberum Zaddach

Trait attributes

Ceriodaphnia lacustris O.F.M.

Ceriodaphnia dubia Richard Ceriodaphnia reticulataJurine

Polyphemus pediculusLinnee

Bosmina longirostris Muller

Chydorus sphaericus Muller

Daphnia ambiguaScourfield Daphnia catawba Coker Daphnia galeata Sars

Alona affinis Leydig Alona rectangula Sars Eubosmina corregoni Baird Bosmina longispina Leydig

Daphnia longispina Mu ller

Chydorus brevilabris Frey

Daphnia magna Straus Daphnia pulex Leydig Daphnia pulicaria Forbes

Sida crystallina Muller

Table 3 Qualitative

Daphnia rosea Sars

observations.

Species

804 A.J. Barnett et al.

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

High [102,119,125,126,128]

Table 4 Qualitative functional traits of copepod species including sources in superscript numbers (see Supplementary material). Trait attributes obtained by generalisation of a group is indicated with a ‘ ’.

High [102,119,125,126]

High [102,119,125,126]

High [102,119,125,126]

High [102,119,125,126]

High [102,119,125,126]

High [102,119,125,126]

High [102,119]

High [102,119]

High [102,119]

High [102,119]

High [102,119]

High [102,119]

High [99,123]

High [96,121]

Selectivity

Current-cruiser [102] Stationary suspension [102] Stationary suspension [102] Stationary suspension [102] Stationary suspension [102] Stationary suspension [102] Stationary suspension [129] Stationary suspension [102]

Raptorial [96,121]

Feeding type

Raptorial [88]

Raptorial [88]

Raptorial [88]

Raptorial [88]

Raptorial [88]

Raptorial [88]

Raptorial [102]

Raptorial [123]

Optimal productivity conditions

High [29]

High [29]

High [29]

High [29]

High [29]

High [29]

High [29]

Omnivore [118] Omnivore [88] Omnivore [88] Omnivore [88,90] Omnivore carnivore [96,121] Omnivore herbivore [122] Omnivore carnivore [123] Omnivore herbivore [102] Omnivore herbivore [88]

Omnivore-carnivore [124] Herbivore [118] Herbivore [118] Herbivore [118] Omnivore [102] Herbivore [118] Omnivore [129] Omnivore [118]

Trophic group

Pelagic [118]

Littoral [127]

Littoral [118]

Pelagic [88]

Pelagic [88]

Pelagic [118]

Pelagic [118]

Pelagic [118]

Pelagic [118]

Pelagic [118]

Pelagic [120]

Pelagic [115]

Pelagic [115]

Pelagic [115]

Pelagic [96]

Habitat

Acanthocyclops vernalis Fischer Cyclops scutifer Sars Cyclops vicinus Ulianine Diacyclops bicuspifatus thomas Forbes Eucyclops speratus Lilljeborg Mesocyclops edax Forbes Tropocyclops prasinus Fischer Thermocyclops crassus Fischer Calanoida Diaptomus minutus Lilljeborg Diaptomus oregonensis Lilljeborg Epischura lacustris Forbes Leptodiaptomus minutus Lilljeborg Leptodiaptomus sicilis Forbes Onychodiaptomus birgei Marsh Skistodiaptomus oregonensis Lilljeborg

Skistodiaptomus reighardi Marsh

robustus Sars

Acanthocyclops

Cyclopoida

Species

Functional diversity of zooplankton communities 805

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

806 A.J. Barnett et al.

be predominantly herbivorous (Table 3), with the exception of Polyphemus. Cyclopoids were all classi- fied as raptorial feeders in the five studies where qualitative data was obtained (Table 4). Other traits, such as preferred food size and productivity condi- tions for optimal growth, were found for some cladoceran species but not for cyclopoids. Data are sparse overall for some variables such as the produc- tivity conditions of optimal growth and qualitative food size.

Functional community dendrogram

The major groups in the taxonomy-based tree of 36 species constructed through simple clustering of species by genus, family, class, etc. (Fig. 1) are similar but not identical to the dendogram generated using length, habitat, trophic group and feeding type (Fig. 2). Five major groups emerge from the FD dendogram (Fig. 2), with one species (Diaphanosoma brachyurum) forming an entire group. Group 1 on Fig. 2 is a mixed group of smaller herbivorous Cladocera and herbivorous (occasionally, herbivor- ous–omnivorous) calanoid and cyclopoid copepods. The upper half of this group is composed of mainly

Ctenopoda Onychopoda Branchiopoda Amomopoda Calanoida Copepoda Cyclopoida
Ctenopoda
Onychopoda
Branchiopoda
Amomopoda
Calanoida
Copepoda
Cyclopoida

Sida crystallina Diaphanosoma brachyurum Holopedium gibberum Polyphemus pediculus Chydorus brevilabris Chydorus sphaericus Alona affinis Bosmina longirostris Bosmina longispina Eubosmina corregoni Daphnia catawba Daphnia ambigua Daphnia rosea Daphnia pulicaria Daphnia pulex Daphnia magna Daphnia longispina Daphnia galeata mendotae Ceriodaphnia reticulata Ceriodaphnia dubia Ceriodaphnia lacustris Skistodiaptomus reighardi Skistodiaptomus oregonensis Onychodiaptomus birgei Leptodiaptomus sicilis Leptodiaptomus minutus Epischura lacustris Mesocyclops edax Eucyclops speratus Thermocyclops crassus Tropocyclops prasinus Cyclops vicinus Cyclops scutifer Diacyclops bicuspifatus thomas Acanthocyclops vernalis Acanthocyclops robustus

Fig. 1 A taxonomic tree of the species used to create the functional dendograms in Figs 2 & 3. Taxonomy is based on Thorp & Covich (2001). Branch lengths are for representation only and do not represent quantitative values.

littoral herbivores (except for Leptodiaptomus sicilis ) while the bottom half of group 1 is mainly pelagic herbivores ( Eucyclops speratus excepted). It appears that herbivorous habit and smaller body size distin- guish group 1. Group 2 contains the remaining copepods, all of which display omnivorous–carnivor- ous tendencies. The cladoceran Polyphemus is also found in this group as it is carnivorous. Group 3 consists of the larger pelagic cladocerans and contains all Daphnia species with the exception of D. pulex and D. magna – both of which are pond/littoral species and fall into group 4 along with the other littoral, large Cladocera. Finally, Diaphanosoma brachyrum forms group 5, as it is an S-filtration species but its much smaller body size appears to warrant its separation from S-filtering Sida. While the dendrogram generated using body length, habitat and trophic group grouped some related taxa together, the dendrogram generated using only body length showed no tendency to group species of a particular genus or family together (Fig. 3).

Discussion

In order to apply a functional perspective to zooplankton communities, we must first establish whether function can be adequately estimated for commonly encountered species. Given the large number of studies on zooplankton life history, mor- phology and physiology over the past few decades, one would expect that a functional classification could be developed. In this review of the literature, we have pieced together the various studies and data therein for common North American species of freshwater cladocerans, and calanoid and cyclopoid copepods. In general, most data were available for cladocerans while there were fewer studies on freshwater cope- pods. This bias reflects the ease of culturing and studying cladocerans, and especially Daphnia species, because of their relatively non-selective filter feeding behaviour and nutritional requirements (Pennak, 1989). Their feeding differs from that of copepods, which actively capture and ingest individual suspen- ded food particles (Koehl & Strickler, 1981). Further, longer generation times and sexual reproduction lead to culturing difficulties which bias studies away from freshwater copepod species. Feeding preferences and rates are important char- acteristics of species that ideally would be included in

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

Fig. 2 A functional dendrogram (FD) generated by hierarchical clustering analysis of the standardised Euclidean distances of each species based on four functional traits (body length, habitat, trophic group and feeding type). A dashed line indicates the split that defines five groupings (numbered) of species as discussed in the text.

Functional diversity of zooplankton communities 807

1 2 3 4 5 43210
1
2
3
4
5
43210

Distance

Onchyodiaptomus birgei Leptodiaptomus sicilis Chydorus brevilabris Chydorus sphaericus

Alona affinis Bosmina longirostris Bosmina longispina Eubosmina corregoni Eucyclops speratus Thermocyclops crassus Tropocyclops prasinus Skistodiaptomus reighardi Skistodiaptomus oregonensis Leptodiaptomus minutus Epischura lacustris Diacyclops bicuspifatus thomas

Polyphemus pediculus Mesocyclops edax Cyclops scutifer Acanthocyclops vernalis Acanthocyclops robustus Cyclops vicinus Daphnia longispina Daphnia catawba Daphnia pulicaria Daphnia galeata mendotae Daphnia rosea Daphnia ambigua

Ceriodaphnia lacustris Ceriodaphnia reticulata Ceriodaphnia dubia Holopedium gibberum

Sida crystallina Daphnia pulex

Daphnia magna Diaphanosoma brachyurum

Fig. 3 A functional dendrogram (FD) generated by hierarchical clustering analysis of the standardised Euclidean distances of each species based solely on body length.

distances of each species based solely on body length. 2.5 2.0 1.5 1.0 0.5 0 Distance
2.5 2.0 1.5 1.0 0.5 0
2.5
2.0
1.5
1.0
0.5
0

Distance

Chydorus sphaericus Bosmina longispina Bosmina longirostris Tropocyclops prasinus Eubosmina corregoni Chydorus brevilabris Leptodiaptomus minutus Diacyclops bicuspifatus thomas Diaphanosoma brachyurum Ceriodaphnia lacustris Polyphemus pediculus Cyclops scutifer Daphnia ambigua Thermocyclops crassus Eucyclops speratus Ceriodaphnia reticulata Ceriodaphnia dubia Alona affinis Mesocyclops edax Acanthocyclops vernalis Skistodiaptomus reighardi Leptodiaptomus sicilis Holopedium gibberum Acanthocyclops robustus Sida crystallina Daphnia magna Daphnia longispina Daphnia catawba Cyclops vicinus Daphnia pulicaria Daphnia pulex Daphnia galeata mendotae Onychodiaptomus birgei Daphnia rosea Epischura lacustris Skistodiaptomus oregonensis

calculations of functional diversity (e.g. FD). We chose to include clearance rate and food size range here, as these were the two most commonly reported standard measures in the literature. The wide range of clearance rates between studies, however, makes it difficult to choose a single clearance rate to include in FD calculations. Clearance rate is known to be affected by a wide range of factors, including food concentra-

tion (Peters & Downing, 1984; Sterner, 1989; Mauch- line, 1998), body size (Peters & Downing, 1984) and temperature (Burns, 1969; Peters & Downing, 1984). We attempted to account for these differences by selecting maximum clearance rate values at one temperature (20 C) and low food concentrations, and we accounted for variations with body size by reporting only values for adult female copepods and

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

808 A.J. Barnett et al.

clearance rate versus body size regressions for cla- docerans. Clearance rate, however, also varies with food type (Bogdan & Gilbert, 1982; Sterner, 1989), food quality (Sterner, 1989; Gulati & DeMott, 1997), predation pressure (Williamson, 1986), hunger (Stem- berger, 1986; DeMott, 1995; Mauchline, 1998), length of experiment (Mauchline, 1998) and the time of day the experiment was run (Schoeneck, Williamson & Stoeckel, 1990). These various factors probably ac- count for the wide range of clearance rates for several species in the trait tables. Despite this considerable variability, an examination of the range of observed values may still provide valuable insights into the relative feeding rates of individual species. Ingestion rates (food quantity consumed per unit time) may also prove to be a useful trait for FD calculations. Many studies of copepods in particular, however, report ingestion rates as the number of food particles consumed per unit time, rather than a rate of mass or carbon ingestion, limiting the comparibility of studies. Ingestion rate also varies with experimental conditions, as noted above for clearance rate, making ingestion rate difficult to include in FD calculations at present. Reported feeding rates in our review have not been adjusted by the mean body size of each species. This is because the goal of an FD calculation is to provide a measure of the overall niche space occupied by a community of organisms. The actual feeding rates of zooplankton are thus necessary to determine what the overall effect of a particular community might be on its resource. In addition, FD calculations are sensitive to highly correlated data (Mason et al., 2005). Thus it would be inappropriate to include both body length and biomass-adjusted feeding rates in a single calculation since individual biomass is most usually calculated through length–weight regressions. As mean body length is the more reliable trait, we have included that in our FD calculations. In terms of feeding preferences, some issues arise when data are collated as we have done. In comparing across studies, it was clear that the characterisation of cyclopoids as carnivores (Sterner, 1989) was not supported. Studies on Tropocyclops prasinus and Meso- cyclops edax reveal that, while both are omnivorous, the former is more herbivorous (Peacock & Smyly, 1983) while the latter is more carnivorous (Confer, 1971), and algae are an important component of the diet of both juvenile and adult cyclopoids (Brandl, 1998). The use of newer food web determination techniques, such as

stable isotopes, may be particularly fruitful for the characterisation of the food sources of zooplankton within lakes of interest in the near future (Bearhop et al., 2004), and such information could be included in an expanded functional classification scheme. The mesh size of filtering apparatus in cladocerans (Table 1) can be useful as a predictor of food selec- tivity, but this may not always be the case. While there is a correlation between uptake efficiency for small food sizes (bacteria) and filtering apparatus size (Brendelberger, 1985), some exceptional species, like Bosmina , do not feed efficiently on food sizes predic- ted from mesh size (DeMott, 1985). Furthermore, ‘soft’ algae (naked, gelatinous flagellates) are more likely to be ingested by filter feeders than ‘hard’ algae (diatoms and dinoflagellates) (DeMott, 1995). An alternative to using quantitative mesh size measurements is to follow the more qualitative groupings suggested by Geller & Mu¨ ller (1981) and DeMott (1985); (1) high efficiency bacteria feeders, (2) low efficiency bacteria feeders and (3) presumptive macrofiltrators. How- ever, we did not use this classification because mesh size is a morphological feature that is more readily measured on preserved zooplankton that would typically be collected and used for FD calculations. For FD calculations, a measure of population growth rate is most informative, but this information is only available at present for some species of cladocerans. Generation time was the most common standard measure of growth available in the litera- ture. These values are also known to be dependent on food quantity and temperature (Hart, 1990; Peterson, 2001; Sarma, Nandini & Gulati, 2005), predation pressure and food quality (Sarma et al., 2005). By selecting the shortest generation times at 20 C, we attempted to account for these variations as much as possible, but discrepancies between studies were still apparent, sometimes making it difficult to select one value of generation time per species for FD calcula- tions. Clutch sizes was also available for many zooplankton species although, per se , clutch size does not reflect the rate of population increase and is of limited value in FD calculations. Egg production rate per female, in number of eggs produced per day or carbon produced per day, would be more useful for FD calculations, although these measures are not as common as clutch size in the literature. Stoichiometric ratios are increasingly being used to characterise the role of species in communities, as

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

they give insight into nutrient cycling (e.g. Elser et al., 2000; Frost et al. , 2006). Given all the caveats and issues with calculation of feeding and growth rates, and given its strong connection with ecosystem effects of species, stoichiometry seems particularly to offer a good measure of functional traits in freshwater zooplankton. Productivity conditions for optimal growth was one qualitative trait for which some information was available. However, this trait should probably be considered together with other habitat characteristics, as there are also important top–down features that will affect zooplankton. DeMott, Edington & Tessier (2004) demonstrated, for example, that the presence of predators, as well as refuges, played a more important role in determining zooplankton species presence and abundance in lakes than did bottom–up nutrient effects. Thus, while these productivity conditions remain useful for determining the food niches of zooplankton species, studies determining the pred- ator or refuge niches of zooplankton would be of benefit. Finally, an important difficulty with the collection of traits from the literature stems from changes in species classification after taxonomic re-evaluation. For example, B. longirostris was considered a common North American species until De Melo & Hebert (1994) analysed allozymic traits and re-analysed morphology with an electron microscope and showed that this species was located only in California. In the rest of the North American distribution, it has been divided into two species; Sinobosmina liederi and Sinobosmina freyi . Functional information on these newly defined species is obviously lacking but prob- ably overlaps significantly with that of B. longirostris . In addition, due to a re-evaluation of the taxonomy of the genus Daphnia, the taxonomic status of Daphnia rosea is now under question (Hebert, 1995) and, thus, data for this species may no longer be applicable in future research. Furthermore, our study was limited to the collection of traits of crustacean zooplankton and did not include rotifers, the latter being an important food web link to microbes, as well as competitors and prey for crustacean zooplankton (Nogrady, Wallace & Snell, 1993). Thus, the determi- nation of similar parameters for rotifers may also be useful for estimates of zooplankton functional diver- sity and the study of the drivers of diversity, niche use and ecosystem function.

Functional diversity of zooplankton communities 809

The decision as to which traits to include in the FD dendogram was therefore a difficult one. Quantitative traits were plagued with the issues of discrepancies between studies, and the lack of data for several groups of species. Furthermore, as functional mea- sures like Petchey & Gaston’s (2002) FD are affected by the use of measures that are correlated, it would be redundant to use correlated measures (Mason et al. , 2005) such as body size and clearance rate predicted from body size regressions. It would be preferable to use only one of these measures at a time. Because of these concerns, we opted to include body length as the only quantitative trait, and habitat, feeding type and trophic group as qualitative traits. The dendrograms in Figs 2 & 3 provide an example of the how a suite of traits can be used to generate a functional dendrogram for a regional species pool. The dendrogram based on four functional traits (Fig. 2) more closely resembled taxonomic groupings, although with some significant exceptions, than did the dendrogram constructed based on body size alone (Fig. 3). This is not surprising as morphological structure is often related to function in zooplankton and, thus, although we argue that taxonomic identi- fication (based largely on morphology) is not suffi- cient, it should still have some relationship with function. The important results lie in the exceptions observed where closely related sister-genera and other related taxa do not cluster together in the functional dendrogram. Group 1 (Fig. 2) was a good example of relatively unrelated taxonomic species that were grouped together in the functional dendrogram. Here a mixture of unrelated herbivores consisting of smaller Cladocera (bosminids, chydorids) and copepods ( Eucyclops , Onchyodiaptomus , Leptodiapto- mus ) were classified together. Similarly, carnivorous species made up another important group. The group most closely related to taxonomic classification was group 3, which contained most of the Daphnia species but also other pelagic cladocerans. Thus, considering a zooplankton community in terms of its ecological functioning should lead to a different diversity calcu- lation than one would obtain using traditional taxo- nomically based diversity indices. Our dendrogram based on four traits provides a starting point for such functional calculations and usefully classifies species based on their ecological roles. The number of traits included can significantly affect the FD measure (Petchey & Gaston, 2002). From

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

810 A.J. Barnett et al.

the dendrogram calculated using body size alone (Fig. 3), the resulting classification does not resemble at all the taxonomic one, and probably does not provide an accurate functional descriptor of the community either. It is therefore evident that the functional dendrogram and estimate of FD can depend on which characters are included in the analysis. Petchey & Gaston (2002) demonstrated the importance of choosing a suite of traits that function- ally characterises the community without overempha- sising functional complementarity; too few traits underestimates the level of complementarity, while too many traits results in a measure that is effectively the same as species richness. We show here the two extremes in what is currently possible to calculate for zooplankton FD based on our review of the literature to date. The ultimate test of which characters to use in combination to calculate FD will be determined by the combination that best represents a particular ecosys- tem function of interest (e.g. total biomass of zooplankton) or most responds to an environmental gradient (such as a gradient of total phosphorus for example, Barnett & Beisner, 2007). Blackburn et al. (2005) conducted such an analysis for the purposes of determining which functional characteristics of a community of invasive predators were most respon- sible for driving bird species extinct on islands. They tested all 2 11 possible combinations of 11 functional traits for predator communities to determine which provided the highest explanatory power for extinc- tions. Such an analysis can also be conducted for zooplankton communities to determine which func- tional traits contribute most to their responses or their function in particular lake ecosystems (Barnett & Beisner, 2007). To obtain Petchey & Gaston’s (2002) FD of a lake community containing only a subset of this regional pool for north-eastern North America, one would need only to sum the branch lengths of those species found in the lake in our dendrogram. Not only do these dendrograms provide an example of how these traits can be used, they show how taxonomic related- ness does not relate directly to function, even when we have had to assume in some cases that species within a genus have similar qualitative characteristics. Large taxonomic differences may not necessarily imply large difference in the ecological roles of species in a community. Therefore, a more functional approach to diversity of lake zooplankton communi-

ties may be very useful, especially when predicting how community function, in addition to structure, may change under future altered conditions.

A practical approach to constructing functional

classifications would combine traits that can be readily

measured. For example, multiple sampling through time of a lake undergoing a seasonal succession provides snapshots of the zooplankton community. Measurements of length can be used to infer feeding rates and food niches to a certain degree, while generation time, population growth rate and age structure can be useful measures of the responses to predation and environmental changes. Stoichiometric values can be applied to the community to determine fluxes of major nutrients through the food web. With the data compiled in this review such an approach can

now begin, with at least a minimum of characteristics. Again, it will be important to test this classification scheme experimentally, to see which combination of traits most accurately captures changes in community function under varied habitat conditions.

In summary, this study provides a synthesis of

available functional traits of zooplankton species based on approximately five decades of laboratory and observational studies. Data were most abundant for cladocerans, particularly Daphnia species, while most lacking for calanoid species. This information is useful to those who wish to estimate functional diversity for North American freshwater zooplankton communities based on taxonomic information of community composition. It also outlines the limits of our current understanding of niche use by zooplank- ton species, and points to a more integrative functional approach to compiling new data on zooplankton that would be beneficial to the ecological study of lakes. The use of functional measurements of diversity in capturing ecological function of aquatic ecosystems has yet to be explored in detail, but work from terrestrial systems shows that such an approach should provide a greater understanding of the mech- anisms which determine community composition and response to major environmental shifts.

Supplementary material

The following supplementary material is available for this article:

Appendix S1. List of references used to generate functional trait matrices.

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

This material is available as part of the online article from: http://www.blackwell-synergy.com/doi/abs/ 10.1111/j.1365-2427.2007.01733.x (This link will take you to the article abstract). Please note: Blackwell Publishing are not respon- sible for the content or functionality of any supple- mentary materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

Acknowledgments

The authors thank NSERC and CFI (Canada), FQRNT (Que´ bec, Canada) for funding to BEB, the foundation of the University of Quebec at Montreal for a post- graduate scholarship to AJB, and a GRIL postdoctoral fellowship to KF. Thanks to M.L. Longhi, M. Maezo and G. Me´ thot for field and lab assistance and to R. Vogt, W. DeMott, A. Hildrew and an anonymous reviewer for helpful comments on the manuscript.

References

Adrian R. (1991) Filtering and feeding rates of cyclopoid copepods feeding on phytoplankton. Hydrobiologia , 210, 217–223. Bady P., Dole´ dec S., Fesl C., Gayraud S., Bacchi M. & Scho¨ ll F. (2005) Use of invertebrate traits for the biomonitoring of European large rivers: the effects of sampling effort on genus richness and functional diversity. Freshwater Biology , 50, 159–173. Barnett A.J. & Beisner B.E. (2007) Zooplankton biodiver- sity and lake tropic state: explanations invoking resource abundance and distribution. Ecology , In press. Bearhop S., Adams C.E., Waldron S., Fuller R.A. & Macleod H. (2004) Determining trophic niche width: a novel approach using stable isotope analysis. Journal of Animal Ecology , 73, 1007–1012. Blackburn T.M., Petchey O.L., Cassey P. & Gaston K.J. (2005) Functional diversity of mammalian predators and extinction in island birds. Ecology , 86,

2916–2923.

Bogdan K.G. & Gilbert J.J. (1982) Seasonal patterns of feeding by natural populations of Keratella, Polyar- thra, and Bosmina: clearance rates, selectivities, and contributions to community grazing. Limnology and Oceanography , 27, 918–934. Brandl Z. (1998) Feeding strategies of planktonic cyclo- poids in lacustrine ecosystems. Journal of Marine Systems , 15, 87–95.

Functional diversity of zooplankton communities 811

Brendelberger H. (1985) Filter mesh-sizes and retentions efficiency for small particles: comparative studies with Cladocera. Ergebnisse der Limnologie , 21, 135–146. Burns C.W. (1969) Relation between filtering rate, tem- perature, and body size in four species of Daphnia . Limnology and Oceanography, 14 , 693–700. Clarke K.R. & Gorley R.N. (2001) PRIMER v5: User Manual/Tutorial . PRIMER-E Ltd., Plymouth, U.K. Confer J.L. (1971) Intrazooplankton predation by Meso- cyclops edax at natural prey densities. Limnology and Oceanography , 4, 663–666. Connell J.H. (1978) Diversity in tropical rain forests and coral reefs. Science, 199 , 1302–1310. De Melo R. & Hebert P.D.N. (1994) A taxonomic reevaluation of North American Bosminidae. Canadian Journal of Zoology , 72, 1808–1825. DeMott W. (1985) Relations between filter mesh-size, feeding mode, and capture efficiency for cladocerans feeding on ultrafine particles. Ergebnisse der Limnologie , 21, 125–134. DeMott W. (1995) Food selection by calanoid copepods in response to between-lake variation in food abundance. Freshwater Biology , 33, 171–180. DeMott W. & Kerfoot W.C. (1982) Competition among cladocerans: nature of the interaction between Bosmina and Daphnia . Ecology 63, 1949–1966. DeMott W., Edington J.R. & Tessier A.J. (2004) Testing zooplankton food limitation across gradients of depth and productivity in small stratified lakes. Limnology and Oceanography , 49, 1408–1416. Diaz S. & Cabido M. (2001) Vive la difference: plant functional diversity matters to ecosystem processes. Trends in Ecology and Evolution , 16, 646–655. Elser J.J., Fagan W.F., Denno R.F. et al. (2000) Nutritional constraints in terrestrial and freshwater food webs. Nature , 408, 578–580. Frank P.W. (1952) A laboratory study of intraspecies and interspecies competition in Daphnia pulicaria (Forbes) and Simocephalus vetulus C.F. Mu¨ ller. Physiological Zoology , 25, 178–204. Frost P.C., Benstead J.P., Cross W.F., Hillebrand H., Larson J.H., Zenopoulos M.A. & Yoshida T. (2006) Threshold elemental ratios of carbon and phosphorus in aquatic consumers. Ecology Letters, 9 , 774–779. Geller W. & Mu¨ ller H. (1981) The filtration apparatus of Cladocera: filter mesh-sizes and their implications on food selectivity. Oecologia , 49, 316–321. Gillooly J.F. (2000) Effect of body size and temperature on generation time in zooplankton. Journal of Plankton Research , 22, 241–251. Gliwicz Z.M. & Pijanowska J. (1989) The role of predation in zooplankton succession. In: Plankton Ecology:

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

812 A.J. Barnett et al.

Succession in Plankton Communities (Ed. U. Sommer), pp. 253–296. Springer-Verlag, New York, U.S.A. Grime J.P. (1977) Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. American Natural- ist , 111, 1169–1194. Gulati R.D. & DeMott W. (1997) The role of food quality for zooplankton: remarks on the state-of-the-art, perspectives and priorities. Freshwater Biology , 38,

753–768.

Hart R.C. (1990) Copepod postembryonic durations:

patterns, conformity, and predictability. The realities of isochronal and equiproportional development, and trends in the copepodid-naupliar duration ratio. Hydrobiologia , 206, 175–206. Hebert P.D.N. (1995) The Daphnia of North America: An Illustrated Fauna . CD-ROM. Department of Zoology, University of Guelph, Ontario. Hooper D.U. & Vitousek P.M. (1997) Effects of plant composition and diversity on ecosystem processes. Science , 277, 1302–1305. Koehl M.A.R. & Strickler J.R. (1981) Copepod feeding currents: food capture at low Reynolds number. Limnology and Oceanography , 26, 1062–1073. Krebs C.J. (1999) Ecological Methodology , 2nd edn. Addison Wesley Longman, Inc, Don Mills Ont, 620 pp. Lavorel S., McIntyre S. & Landsberg T.D.A. (1997) Plant functional classifications: from general groups to specific groups based on response to disturbance. Trends in Ecology and Evolution , 12, 474–476. Lynch M. (1980) The evolution of cladoceran life histor- ies. The Quarterly Review of Biology , 55, 23–42. Maier G. (1994) Patterns of life history among cyclopoid copepods of central Europe. Freshwater Biology , 31,

77–86.

Mason N.W.H., Mouillot D., Lee W.G. & Wilson J.B. (2005) Functional richness, functional evenness and functional divergence: the primary components of functional diversity. Oikos, 111 , 112–118. Mauchline J. (1998) The biology of calanoid copepods. Advances in Marine Biology , 33, 1–710. McCauley E. (1984) The estimation of the abundance and biomass of zooplankton in samples. In: A Manual on Methods for the Assessment of Secondary Productivity in Fresh Waters (Eds J.A. Downing & F.H. Rigler) pp. 228– 265. Blackwell Scientific Publications, Oxford. McCauley E. & Briand F. (1979) Zooplankton grazing and phytoplankton species richness: field tests of the predation hypothesis. Limnology and Oceanography , 24,

243–252.

McNaughton S.J. (1977) Diversity and stability of ecological communities: a comment on the role of

empiricism in ecology. The American Naturalist, 111,

515–525.

Mouillot D., Mason W.H.N., Dumay O. & Wilson J.B. (2005) Functional regularity: a neglected aspect of functional diversity. Oecologia , 142, 353–359. Nogrady T., Wallace R.L. & Snell T.L. (1993) Rotifera . SPB Academic Publishing, The Hague, The Netherlands. Norberg J. (2004) Biodiversity and ecosystem function- ing: a complex adaptive systems approach. Limnology and Oceanography , 49, 1269–1277. Peacock A. & Smyly W.J.P. (1983) Experimental studies on the factors limiting Tropocyclops prasinus (Fischer) 1860 in an oligotrophic lake. Canadian Journal of Zoology , 61, 250–265. Pennak R.W. (1989) Fresh-water Invertebrates of the United States: Protozoa to Mollusca . John Wiley & Sons, New York, U.S.A. Petchey O.L. & Gaston K.J. (2002) Functional diversity (FD), species richness and community composition. Ecology Letters, 5, 402–411. Peters R.H. & Downing J.A. (1984) Empirical analysis of zooplankton filtering and feeding rates. Limnology and Oceanography , 29 , 763–782. Peterson W.T. (2001) Patterns in stage duration and development among marine and freshwater calanoid and cyclopoid copepods: a review of rules, physiolo- gical constraints, and evolutionary significance. Hydrobiologia , 453/454, 91–105. Rabette C., Thouvenot A. & Lair N. (1998) Laboratory experiments on trophic relationships and remote detection between two ciliates Cyclops vicinus . Hydro- biologia 373/374, 157–167. Reynolds C.S. (1980) Phytoplankton assemblages and their periodicity in stratifying lake systems. Holarctic Ecology , 3, 141–159. Reynolds C.S., Huszar V., Kruk C., Naselli-Flores L. & Melo S. (2002) Towards a functional classification of the freshwater phytoplankton. Journal of Plankton Research , 24, 417–428. Santer B. (1993) Potential importance of algae in the diet of adult Cyclops vicinus. Freshwater Biology, 30, 269–278. Sarma S.S.S., Nandini S. & Gulati R.D. (2005) Life history strategies of cladocerans: comparisons of tropical and temperate taxa. Hydrobiologia, 542, 315–333. Schoeneck L.J., Williamson C.E. & Stoeckel M.E. (1990) Diel periodicity and selectivity in the feeding rate of the predatory copepod Mesocyclops edax . Journal of Plankton Research , 12, 29–40. Statzner B., Bis B., Dole´ dec S. & Usseglio-Polatera P. (2001) Perspectives for biomonitoring at large spatial scales: a unified measure for the functional composition of invertebrate communities in European running waters. Basic and Applied Ecology , 2, 73–85.

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813

Stemberger R.S. (1986) The effects of food deprivation, prey density and volume on clearance rates and ingestion rates of Diacylops thomasi. Journal of Plank- ton Research , 8, 243–251. Sterner R.W. (1989) The role of grazers in phytoplankton succession. In: Plankton Ecology: Succession in Plankton Communities (Ed. U. Sommer), pp. 107–170. Springer- Verlag, New York. Symstad A.J., Siemann E. & Haarstad J. (2000) An experimental test of the effects of plant functional group diversity on arthropod diversity. Oikos , 89,

243–253.

Thorp J.H. & Covich A.P. (2001) Ecology and Classification of North American Freshwater Invertebrates , 2nd edn. Academic Press, San Diego. Tilman D., Knops J., Wedin D., Reich P., Ritchie P. & Siemann E. (1997) The influence of functional diversity and composition on ecosystem processes. Science , 277,

1300–1302.

Walker B.H. (1991) Biodiversity and ecological redund- ancy. Conservation Biology , 6, 18–23. Walker B.H. & Langridge J.L. (2002) Measuring func- tional diversity in plant communities with mixed life forms: a problem of hard and soft attributes. Ecosys- tems , 5 , 529–538.

Functional diversity of zooplankton communities 813

Walker B., Kinzig A. & Langridge J. (1999) Plant attribute diversity, resilience, and ecosystem function: the nature and significance of dominant and minor species. Ecosystems , 2, 95–113. Weithoff G. (2003) The concept of ‘plant functional types’ and ‘functional diversity’ in lake phytoplankton – new understanding of phytoplankton ecology? Freshwater Biology , 48, 1669–1675. Williamson C.E. (1986) The swimming and feeding behavior of Mesocyclops . Hydrobiologia , 13, 11–19. Witty L.M. (2004) Practical Guide to Identifying Freshwater Crustacean Zooplankton . Cooperative Freshwater Ecol- ogy Unit, Sudbury, Ontario. Wyngaard G.A., Rasch E.M., Manning N.M., Gasser K. & Domangue R. (2005) The relationship between genome size, development rate, and body size in copepods. Hydrobiologia , 532, 123–137. Yachi S. & Loreau M. (1999) Biodiversity and ecosystem productivity in a fluctuating environment: The insur- ance hypothesis. Proceedings of the National Academy of Science of the United States of America, 96 , 1463–1468.

(Manuscript accepted 5 January 2007)

2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology, 52, 796–813