Latitudinal variation in nematode diversity and ecological roles along the Chinese coast

Abstract Aim To test changes in the phylogenetic relatedness, niche breadth, and life‐history strategies of nematodes along a latitudinal gradient. Location Sixteen wetland locations along the Pacific coast of China, from 20°N to 40°N. Methods Linear regression was used to relate nematode phylogenetic relatedness (average taxonomic distinctness (AvTD) and average phylogenetic diversity [AvPD]), life‐history group (based on “c‐p” colonizer–persister group classification), and dietary specificity (based on guild classification of feeding selectivity) to latitude. Results Wetland nematode taxonomic diversity (richness and Shannon diversity indices) decreased with increasing latitude along the Chinese coast. Phylogenetic diversity indices (AvTD and AvPD) significantly increased with increasing latitude. This indicates that at lower latitudes, species within the nematode community were more closely related. With increasing latitude, the nematode relative richness and abundance decreased for selective deposit feeders but increased for nonselective deposit feeders. The proportion of general opportunists decreased with increasing latitude, but persisters showed the opposite trend. The annual temperature range and the pH of sediments were more important than vegetation type in structuring nematode communities. Main conclusion Nematode niche breadth was narrower at lower latitudes with respect to dietary specificity. Higher latitudes with a more variable climate favor r over K life‐history strategists. Nematode communities at lower latitudes contained more closely related species.

also to account for changes in ecological niches. For instance, Krasnov, Shenbrot, Khokhlova, Mouillot, and Poulin (2008) found that the niche breadth of parasitic fleas increased at higher latitudes, both in terms of host specificity and geographic range size. Based on metaanalysis, González-Bergonzoni, Meerhoff, Teizeira-de Mello, Baattrup-Pedersen, and Jeppesen (2012) reported a global pattern in the relative richness of omnivorous fish species with decreasing latitude.
The clutch size of birds was reported to decrease toward the equator in both hemispheres (Cardillo, 2002;Hille & Cooper, 2015). Revealing the latitudinal pattern of ecological niches and roles is important for understanding different ecosystem processes at different latitudes (Vázquez & Stevens, 2004).
Ecological traits of organisms are often linked to phylogenetic history, as closely related species tend to share similar resource requirements and responses to the environment (Wiens et al., 2010). Many low-latitude clades lack necessary ecological adaptations to survive harsher environments (e.g., low temperature) and cannot cross ecophysiological barriers to unfavorable regions (Ricklefs, 2006). Despite exceptions in polar regions (Adams et al., 2006), coexisting species in a community are often found to be more phylogenetically related at high latitudes where the environment is more dynamic (Qian, Zhang, Zhang, & Wang, 2013). These patterns are mostly proposed for several taxa including plant, bird, and mammal species (Cardillo, Orme, & Owens, 2005;Qian et al., 2013;Safi et al., 2011), and it is still unclear for most taxonomic groups and studies on changes in phylogenetics are far behind those on changes in species richness with latitudinal gradients.
Soil biota consists of a wide range of life forms, which contribute to many essential ecological functions, including decomposition, carbon and nutrient cycling, disease suppression, and regulation of plant growth and primary productivity (Bardgett, Yeates, & Anderson, 2005). Understanding latitudinal patterns in the diversity, ecological niche, and phylogenetic composition of soil biota is important for exploring global patterns of ecosystem functioning (Brussaard et al., 2012). However, the latitudinal patterns of ecological characteristics for belowground biota are more poorly understood and may be different from that of aboveground organisms (Bardgett & van der Putten, 2014; Bardgett et al., 2005;Decaëns, 2010;Wu, Ayres, Bardgett, Wall, & Garey, 2011).
Among soil biota, nematodes hold a central position due to their high abundance and pivotal roles in belowground food webs and ecosystem functioning (Neher, 2010). Reports describing the variations in ecological function and phylogenetic composition of nematode communities across latitudinal gradients are scarce. This study aimed to explore the ecological and evolutionary aspects of nematode diversity distribution across a latitudinal range along the Chinese coast. Using a standardized sampling approach, we collected nematodes ( Figure 1) from 16 coastal wetland sites, spanning latitudes from 20°N to 40°N.
The relationships of nematode ecological groups (groups with different dietary habits, life-history strategies, and phylogenetic relatedness) and environmental factors were analyzed. Based on the literature describing these relationships for aboveground taxa, we tested the following hypotheses: (1) nematode taxonomic richness and diversity decrease with increasing latitude, (2) species within communities are more phylogenetically clustered at high latitudes, (3) nematode niche breadth is wider toward high latitudes, and therefore, more dietary generalists are found at high latitudes, and (4) high latitudes are more favorable to large-bodied r-strategists compared with K-strategists.

| Sampling locations and climate data
Sampling was conducted at 16 locations along a latitude ranging from 20.64°N to 40.88°N along the Pacific coast of China (Figure 2).
At each location, several (1-7) marsh or mangrove wetlands spaced more than 500 m apart were selected (Table 1) At each of the 53 wetlands from 16 sampling localities, five sediment cores were collected at 2-m intervals along an 8-m transect. Samples were combined together to form a composite sample. Sediment samples were homogenized and then split into two parts. About 150 g of sediment was fixed in 4% hot formalin for nematode analysis. The remainder of the sample was air-dried and analyzed for sediment grain size composition, total carbon, total nitrogen, and pH. Sediment grain size was measured using a vibrating sieve shaker (FRITSCH analysette 3, Germany). Total carbon and total nitrogen was measured using a NC Soil Analyzer (Flash EA 11121 Series, Thermo Finnigan, Italy). The pH value was measured using a pH meter (IQ150, USA).

| Nematodes
Nematodes were extracted by flotation in Ludox TM in the laboratory. After counting the total number of nematode individuals, about 120 specimens per sample were randomly selected for identification. Nematodes were identified to morphospecies (morphologically distinct taxa) level following Goodey (1963), Jairajpuri and Ahmad (1992), and Warwick, Platt, and Smoerfield (1998). For simplicity, the morphospecies were referred as "species" hereafter. The Shannon-Weaver diversity index H′ = Σp i ln p i , where p i is the proportion of individuals in the ith taxon, was calculated at the species and genus level.
Based on characteristics of the buccal cavity, nematodes were identified as plant feeders or free-living nematodes. Free-living nematodes were further classified into four groups based on feeding types (Wieser, 1953): (1) selective deposit feeders which have no buccal cavity or only a narrow tubular buccal cavity, (2) nonselective deposit feeders which have a large buccal cavity without teeth, (3) epistrate or diatom feeders which have a buccal cavity armed with teeth, and (4) predators or omnivores which have large teeth or jaws.
Based on life history of nematodes at the family level, nematodes were designated a colonizer-persister (c-p) value ranging from 1 to 5 (Bongers, 1990). Nematodes with low c-p values have a short generation time and high fecundity and are comparable to r-strategies in the loose sense (Bongers, 1990). On the other hand, nematodes with high c-p values have a long generation time, large body size and low fecundity and are sensitive to disturbance and can be considered as K-strategies (sensu lato) (Bongers, 1990). Nematodes were classified into one of three categories: enrichment opportunists (c-p 1), general opportunists (c-p 2), and persisters (c-p 3-5) (De Goede, Bongers, & Ettema, 1993).
The proportion of species representing "relative species richness" and the proportion of individuals representing "relative abundance" for each nematode ecological group were calculated. The proportion of species was calculated as S i /S t , where S i is the species richness of a certain ecological group within a community, and S t is the total species richness of the whole community. The proportion of individuals of each group was calculated as I i /I t , where I i is the individual number of a certain ecological group within a community, and I t is the total number of individuals in the whole community.
To assess the phylogenetic composition of a certain nematode community, two indices, average taxonomic distinctness (AvTD) and average phylogenetic diversity (AvPD), were selected. AvTD is a distance-based index calculated using pairwise distance. A higher value of AvTD or AvPD indicates a larger phylogenetic relationship among taxa. According to Schweiger, Klotz, Durka, and Kuehn (2008), AvTD is suggested as the best index to compare independent communities because it is unbiased by species richness. AvPD represents a distance-based index and uses a minimum spanning path to reflect phylogenetic skewness.
Average taxonomic distinctness (AvTD) (Clarke & Warwick, 1998) is a measure of the average degree to which species in an assemblage are related to each other. AvTD = [Σ Σi<j ωij]/[s(s-1)/2], where s is the number of species present, the double summation is over {i = 1, … s; j = 1, … s, such that i < j}, and ωij is the "distinctness weight" given to the path length linking species i and j in the hierarchical classification. The taxonomic levels used in this study are species, genus, family, order, and class, according to the classification described by Lorenzen (1981).
Values of AvTD are based on equal step lengths between the above five taxonomic levels. Thus, the weighting between taxonomic levels for different species of the same genus is ω = 20, for species in the same family but different genera ω = 40, for species in the same order but different families ω = 60, and for species in different classes ω = 100.
Average phylogenetic diversity (AvPD) (Clarke & Warwick, 2001) is a measure of the average amount of phylogenetic diversity (branch length) contributed by randomly chosen species to the total phylogenetic diversity (PD). AvPD = PD/s = Σn i /s, where s is the number of species present, and n i is the number of i nodes within the minimum spanning path.

| Data analyses
Linear regression was used to determine to what degree nematode richness, H′ diversity indices, feeding type, and colonizer-persister (c-

| RESULTS
A total of 166 nematode morphospecies were collected from 16 locations belonging to 112 genera, 44 families, 7 orders, and 2 classes (Appendix S1). Richness and H′ diversity indices of nematodes at both the species level and genus level significantly decreased with increasing latitude (Figure 3). This indicates that southern localities had a greater diversity of nematodes compared with northern locations.
Phylogenetic diversity indices (AvTD and AvPD) significantly increased with increasing latitude (Figure 4). This indicates that the nematode communities at the lower latitudinal locations contained more closely related species.
The proportion of selective deposit feeders was negatively related to increasing latitude, for both species and individuals (Figure 5a,c).
However, the proportion of nonselective deposit feeders showed a weak positive relationship with increasing latitude for species ( Figure 5b). The proportion of individual nonselective deposit feeders also showed a positive relationship with increasing latitude that was statistically marginally significant (r = .266, p = .054, data not shown).
Significant latitudinal gradients were not detected for other trophic groups including plant feeders, epistrate or diatom feeders, and predators or omnivores (data not shown).
The proportion of nematode species and individuals that were classified as general opportunists (c-p 2) showed a significant increase with increasing latitude. However, the proportion of species and individuals that were classified as persisters (c-p 3-5) showed the opposite trend ( Figure 6). No significant relationships between nematodes classified as enrichment opportunists (c-p 1) and latitude were observed (data not shown).
The climatic parameters annual temperature (AT) and annual temperature range (ATR) were the most frequently selected parameters in models developed to investigate variables that may be associated with nematode distribution (Appendix S2). The normalized difference vegetation index (NDVI), which represents vegetation and soil condition measures, was relatively a weak predictor. Correlation coefficient analyses showed that the taxonomic diversity of nematodes (richness and H′ at both the species level and generic level) were positively correlated with AT, but negatively correlated with ATR and pH ( Table 2).
The phylogenetic diversity indices (AvTD and AvPD) were both negatively correlated with AT, but positively correlated with ATR and pH.
The proportion of nematodes that were classified as selective deposit feeders and as persisters (c-p 3-5) was positively associated with AT and precipitation and negatively associated with ATR and pH (Table 2).
Conversely, the proportion of nonselective deposit feeders and general opportunists (c-p 2) exhibited the opposite pattern (Table 2).
Based on the nonmetric MDS plots, the 16 sampling sites were

| Nematode richness, diversity, and community structure
Studies based on synthesized data of nematode diversity mainly yielded two different findings. Some studies concluded that nematode diversity increased with increasing latitude (e.g., Boucher, 1990;Boucher & Lambshead, 1995;Procter, 1984), and others concluded that there were no clear trends in the diversity of nematodes with changing latitude (e.g., Boag & Yeates, 1998;Fonseca & Netto, 2015;Mokievsky & Azovsky, 2002). These contradictory findings may result from unbalanced sampling effort and taxonomic resolution on nematodes from different latitudes (Lee & Riveros, 2012). Therefore, a full understanding of latitudinal patterns of nematode distribution on a global scale may require a more complete taxonomic knowledge and molecular information of this highly diversified group (Wu et al., 2011).
However, their findings may be confounded by sampling depth (Rex, Stuart, & Etter, 2001). Gobin and Warwick (2006) did not find a clear trend in nematode diversity with change in latitude. Given that they collected nematodes using artificial collectors at only four geographic locations, their results may differ from studies that directly sample the soil or sediment. Therefore, in terms of diversity per unit area, the pattern of higher species richness at lower latitudes may hold true for nematodes, similar to the majority of other organisms. The diversity of ants was positively correlated with annual temperature and negatively correlated with temperature range (Dunn et al., 2009). Similarly, we found that nematode richness and diversity in coastal wetlands were positively correlated with annual temperature and negatively correlated with temperature range. Fonseca and Netto (2015) discovered that nematode species composition differed significantly between estuaries with and without mangroves. They speculated that differences in root systems and leaf decomposition processes between mangrove estuaries and salt marshes shaped the different nematode communities. Other studies also have reported the importance of vegetation type on nematode community composition (Nielsen et al., 2014). However, in our study, F I G U R E 5 Relationship between latitude and different nematode feeding groups for (a) relative species richness of selective deposit feeders, (b) relative species richness of nonselective deposit feeders, and (c) relative abundance of selective deposit feeders. n = 53 showed that at latitudes of less than 26°N, nematode communities in mangrove habitats and P. australis marshes could not be distinguished.
At more northern locations (latitudes of 28°N or higher), nematode communities from marshes dominated by P. australis and S. salsa also overlapped. Based on BIO-ENV analyses, our study showed that climate variables, such as annual temperature range (ATR) and sediment properties such as pH, were more important than the dominant vegetation type.

| Nematode phylogenetic diversity
Negative relationships between increasing latitude and phylogenetic diversity have been reported for angiosperm plants (Qian et al., 2013), mammals (Safi et al., 2011), etc. Phylogenetically closely related species tend to share similar habitat requirements and thus are less likely to coexist if the community is structured by competition (Graham, Parra, Rahbek, & Mcguire, 2009;Ulrich & Fattorini, 2013). However, our results showed a positive relationship between nematode phylogenetic diversity (both average taxonomic distinctness AvTD and average phylogenetic diversity AvPD) and increasing latitude. This pattern is probably because small organisms or soil biota are weakly structured by competition (Decaëns, 2010;Ulrich & Fattorini, 2013).
Species living in harsh environments are reported to be more phylogenetically clustered (Graham et al., 2009). This may be because few clades can across ecophysiological barriers from benign environments to harsher ones (e.g., colder or with more disturbance) (Helmus et al., 2010). However, our results indicate that nematodes are phylogenetically more dispersed in colder regions with more variable climates (Table 2). These inconsistencies may result from other biotic interactions such as facilitation or the mobility of organisms that can also affect phylogenetic structure along a stressful habitat gradient (Graham et al., 2009). The exceptions in Antarctica (taxa formed from a wide phylogenetic base) also suggest that it may be related to the possible endemicity and broad dispersal of small-sized soil organisms (Adams et al., 2006(Adams et al., , 2007. Pattern of nematode community assembly, speciation, extinction with latitudes needs further investigation by utilizing phylogenetic information.

| Nematode dietary breadth
A positive relationship between niche breadth and latitude has long been assumed (Morin & Chuine, 2006). According to the pervasive "climate variability hypothesis," greater fluctuations in climate at high latitudes cause more variability in resource availability in time and space and thus favor wider niches (Cardillo, 2002;Slove & Janz, 2010). By studying the feeding plasticity of birds, Simon, Diamond, and Schwab (2003) found that Canadian southern forests were dominated by specialists and northern forests by generalists. Krasnov et al. (2008) found that in terms of host specificity, the niche breadth of parasitic fleas on mammals increased with latitude. Dietary specialization determines an organism's resource base and constitutes an important part of niche breadth. In this study, we analyzed selective and nonselective groups of deposit feeding nematodes separately.
Selective-feeding nematodes are considered to be more specialized in T A B L E 2 Correlation coefficients of taxonomic richness and diversity, phylogenetic diversity, relative abundance, and richness of different feeding groups and colonizer-persister groups with climatic parameters and physiochemical parameters of soil (AT: annual temperature; ATR: annual temperature range; AP: annual precipitation; NDVI: yearly normalize difference vegetation index; TN %: total nitrogen; TC %: total carbon; sand %; pH). Significance levels are marked by asterisk (*p < .05, **p < .01,***p < .001)

| Nematodes with different life-history strategy
Latitude-related variability in the environment not only influences resource use (e.g., dietary specialization), but may also alter the responses of organisms to the environment. Although the ecological traits of K-and r-strategists were not directly measured in this study, the classification of nematodes along the colonizers (c) to persisters (p) spectrum reflects some typical characteristics of life-history responses. We found that the relative abundance and richness of general opportunists (c-p 2) increased with increasing latitude, lower annual temperature, and a higher temperature range. Nematodes that were classified as persisters (c-p 3-5) and showed typical K-selected characteristics decreased in relative abundance and richness with increasing latitude. Studies of other taxonomic groups support our findings that K-strategists may be more common at lower latitudes.
For instance, birds from low latitudes have been found to possess Kstrategy characteristics such as smaller clutch size, large eggs, reduced mortality, and higher survival (Cardillo, 2002;Hille & Cooper, 2015).
For plants, a trend of increasing seed size toward the tropics has been reported (Morin & Chuine, 2006). However, the generalization that more K-strategists occur at lower latitudes is still under debate (Auer & King, 2014). High-latitude gammaridean amphipod species were characterized by large body size, long generation, few broods, large embryos which showed K-selected life-history strategy (Sainte-Marie, 1991). In Europe, higher-latitude populations of freshwater fish have been found grew more slowly, matured later, had a longer life span than lower-latitude populations (Blanck & Lamouroux, 2007). Highlatitude coastal reefs are found to be typified by stress-tolerant generalist coral species (Sommer, Harrison, Beger, & Pandolfi, 2014). For nematode, the highly abundant nematode species Scottnema lindsayae in Antarctic continent soils has a long lifecycle (218 days) and development time, low fecundity, which more closely resembles a K-selective life-history strategy (Adams et al., 2007). The pattern observed in our study may be explained as high local climate variability at higher latitudes benefits r-strategists, which may recover more easily from disturbance (Gaston, 2000;Hille & Cooper, 2015). However, the true underlying mechanism needs further investigation.
Both large-scale climatic variables and small-scale soil parameters could be determinants in structuring nematode community. In addition to the variables including annual temperature, annual temperature range, and soil pH which were significantly associated with nematode diversity, annual precipitation and total soil nitrogen were also found relating to the relative abundance and species richness of nematode ecological groups. This suggests that a wider range of factors may be associated with variations among nematode ecological groups compared with the number of factors that influence nematode species diversity.

| CONCLUSION
Information on the geographic distribution of nematodes is scarce for the Asian region. This is the first study to examine the association between nematode diversity and latitude with respect to dietary specialization and life-history strategies. In relation to our initially proposed hypotheses, we conclude the following: (1) The relative richness and abundance of nematodes that were dietary specialists tend to be greater at lower latitudes.
(2) Nematodes with r life histories were favored over K life-history strategists at higher latitudes, where the climate was more variable.
(3) Nematode species within a community tended to be more closely related in phylogeny at lower latitudes, which may imply that these communities were weakly structured by competition. (4) Wetland nematode diversity decreased with increasing latitude from 20°N to 40°N along the Chinese coast, and annual temperature range and soil pH were more important than vegetation type in influencing nematode community structure.

ACKNOWLEDGMENTS
This study was financially supported by Ministry of Science and