- Email: [email protected]
Dissecting the plant–insect diversity relationship in the Cape
Molecular Phylogenetics and Evolution 51 (2009) 94–99
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev
Dissecting the plant–insect diversity relationship in the Cape Sß erban Prochesß a,b,c,*, Félix Forest d, Ruan Veldtman b,e, Steven L. Chown b, Richard M. Cowling a, Steven D. Johnson f, David M. Richardson b, Vincent Savolainen d,g a
Botany Department, Nelson Mandela Metropolitan University, P.O. Box 77000, Port Elizabeth 6031, South Africa Centre for Invasion Biology, Department of Botany and Zoology, Stellenbosch University, Private Bag X1, Matieland 7602, South Africa c School of Environmental Sciences, University of KwaZulu-Natal, Private Bag X54001, Durban 4000, South Africa d Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3DS, United Kingdom e South African National Biodiversity Institute, Kirstenbosch Research Centre, Private Bag X7, Claremont 7735, South Africa f School of Biological and Conservation Sciences, University of KwaZulu-Natal, Private Bag X01, Scottsville 3209, South Africa g Imperial College London, Silwood Park Campus, Buckhurst Road, Ascot, Berkshire, SL5 7PY, United Kingdom b
a r t i c l e
i n f o
Article history: Received 3 January 2008 Revised 23 May 2008 Accepted 27 May 2008 Available online 7 June 2008 Keywords: Accumulation curves Cape Floristic Region Diversity correlations Phylogenetic diversity Plant–insect interactions Spatial scale Taxonomic rank
a b s t r a c t It has been argued that insect diversity in the Cape is disproportionately low, considering the unusually high plant diversity in this region. Recent studies have shown that this is not the case, but the precise mechanisms linking plant diversity and insect diversity in the Cape are still poorly understood. Here we use a dated genus-level phylogenetic tree of the Cape plants to assess how plant phylogenetic diversity compares with taxonomic diversity at various levels in predicting insect diversity. We find that plant phylogenetic diversity (PD) is a better predictor of insect species diversity that plant species diversity, but the number of plant genera is overall as good a predictor as PD, and much easier to calculate. The relationship is strongest between biomes, suggesting that the relationship between plant diversity and insect diversity is to a large extent indirect, both variables being driven by the same abiotic factors and possibly by common diversification, immigration and extinction histories. However, a direct relationship between plant diversity and insect diversity can be detected at fine scales, at least within certain biomes. Diversity accumulation curves also indicate that the way plant phylogenetic diversity and the number of plant genera increase over spatial scales is most similar to that for insect species; plant species show a greater increase at large spatial scales due to high numbers of local endemics. Ó 2008 Elsevier Inc. All rights reserved.
1. Introduction An important step in the estimation of insect diversity worldwide was the recent finding that plant phylogenetic diversity is a good predictor of herbivorous insect species diversity (Novotny´ et al., 2006). However, this finding was based on phylogenetic trees including a very limited number of plant species. The comprehensive phylogenetic tree for Cape angiosperm genera (Forest et al., 2007) opens up new possibilities of analyzing the influence of plant phylogenetic diversity on insect diversity. In particular, the availability of a spatially explicit data set of plant and insect occurrences (Prochesß and Cowling, 2006) makes it possible to explore the influence of spatial scale on the relationship between insect diversity and plant diversity—while measuring the latter in terms of species, higher taxa, or cumulative phylogenetic distance. There are several aspects that can be clarified by combining these data. It is not clear whether the plant–insect diversity rela-
* Corresponding author. Address: School of Environmental Sciences, University of KwaZulu-Natal, Private Bag X54001, Durban 4000, South Africa. Fax: +27 31 2607317. E-mail address: [email protected] (S ß . Prochesß). 1055-7903/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2008.05.040
tionship holds for non-herbivorous insects and across different spatial scales, or whether plant phylogenetic diversity should be a predictor of any value for insects other than herbivores. Indeed, herbivores represent less than half of all insect species across a variety of ecosystems (Stork, 1987; Krüger and McGavin, 2001), and the diversity relationships between various insect feeding guilds are intricate (Stireman and Singer, 2003). The concordance between plant and insect diversity may hold across feeding guilds if plant and insect diversity respond similarly to environmental factors such as climate and topography (Hawkins and Porter, 2003). If this is the case, predictions of insect diversity patterns from plant diversity data would only be accurate across spatial scales that encompass substantial variation in climate or topography (Gering and Crist, 2002; Finlay et al., 2006). Here, we try to answer some of these questions by examining how the plant–insect diversity relationship varies with spatial scale, taxonomic rank and between insect feeding guilds. Our data come from one of the most plant–rich terrestrial areas of the world, i.e. four biomes (fynbos, grassland, Albany thicket and semi-arid Nama-karoo) across two global biodiversity hotspots in South Africa (the Cape Floristic Region and Maputaland–Pondoland–Albany; Mittermeier et al., 2004), although the magnitude of insect
Sß. Prochesß et al. / Molecular Phylogenetics and Evolution 51 (2009) 94–99
diversity in this region is disputed (Johnson, 1992; Wright and Samways, 1998; Prochesß and Cowling, 2006). 2. Methods The analyses are based on plant surveys and insect collections in 10- by 10-m plots. Insects were collected by exhaustively sweep-netting the vegetation. (Each plant or branch higher than 10 cm above ground was vigorously hit with the net at least once. All plants in the plots were sweep-netted, including shrubs and small trees; there were no large trees in any of the plots.) Eight plots were considered per locality, representing four co-linear sets of two adjacent plots, separated by 100 m, 1 km and 100 m. One locality for each biome was sampled in the Baviaanskloof Conservation Area in the Eastern Cape; additionally, for a second locality, fynbos plots were sampled in the Kogelberg Biosphere Reserve, Western Cape; grassland plots near Dordrecht in the Eastern Cape Drakensberg; Albany thicket plots in the Aloes Nature Reserve in Port Elizabeth, Eastern Cape; and Nama-karoo near Beaufort West, Western Cape. Insect collections were preformed during the season of maximum insect activity at each site: October in the Baviaanskloof and Kogelberg, February in the Drakensberg; December in Port Elizabeth; and April near Beaufort West. This sampling programme gave five spatial scales for analysis: 10 m (plot scale; n = 64); 20 m (pooled data for adjacent plots; n = 32); 100 m (pooled data for plots separated by distances of 100 m; n = 16); 1 km (pooled data for all plots in one locality; n = 8); and biome-wide (pooled data for the two localities in the same biome; n = 4). Plants were identified to species level, and insects assigned to morphospecies. In total, we listed 440 plant species and 636 insect morphospecies (59% phytophagous, 25% predaceous/parasitic and 16% detritivorous, on the basis of predominant feeding habits at family level). Both groups were assigned to higher ranks according to recent taxonomic treatments (APG, 2003; Grimaldi and Engel, 2005). In the case of plants, unplaced taxa that represent sister lineages to orders and subclasses were given order and subclass rank, respectively. Phylogenetic diversity (PD) was calculated from a tree including most angiosperm genera in the Cape (Forest et al., 2007). Sequences were not available for twenty-six of the 228 genera in our plots, and these were deleted from plant lists in calculating PD. This is unlikely to greatly affect the results, as most of them represented large families already present in every plot. The PD lost by not including intra-generic branch lengths is also likely to be minimal (Prochesß et al., 2006). At each spatial scale, each log-transformed insect diversity measure was regressed against each log-transformed plant diversity measure and the r2 fit of a linear model was extracted, as was the p value. The false detection rate method was used to correct for multiple comparisons (García, 2004). The results were found to be qualitatively consistent if using robust linear regressions or generalized linear models with Poisson errors. Accumulation curves for plant taxa, plant phylogenetic diversity, and insect species across spatial scales were drawn based on averaged values at each spatial scale. All analyses were implemented in the package R 2.3.1. (R Development Core Team; http://www.R-project.org). The robust linear regressions used function lmrob (library ‘robustbase’ version 0.2–3). 3. Results The plant–insect diversity relationship varied notably between biomes. In fynbos there was a consistent positive relationship between plant diversity and insect diversity, plant phylogenetic diversity being a better predictor of insect species diversity (as compared to the number of plant species). In grassland, the relationship be-
95
tween plant diversity and insect diversity was actually a negative one, at least at fine spatial scales. In Albany thicket, the relationship was positive, and plant species diversity and plant phylogenetic diversity performed equally well in predicting insect species diversity at fine scales, but, at the 1 km scale, the locality richer in plants was poorer in insects. There was no clear relationship between plant diversity and insect diversity in Nama-karoo (cf. Fig. 1). However, when analyzing data from all biomes together, the insect-plant diversity relationship was consistently positive across spatial scales and taxonomic ranks. The strength of this relationship varied as follows. First, the amount of variation in insect richness explained by plant richness increased with spatial scale (Fig. 2). This increase was most notable at large spatial scales when the predictor was the number of plant species, the number of plant genera, or plant phylogenetic diversity. When plant families, orders and subclasses were considered, there was a major increase in variation explained from the 10 m scale to the 20 m scale, and again from the 1 km scale to the biome-wide scale, but a slight decrease from 20 m to 1 km. For most taxonomic ranks, over 75% of the variation was explained at the biome scale, but this high percentage was largely due to the inclusion of the Nama-karoo biome, where both plant and insect diversity are much lower than in any of the other three biomes. Second, the plant–insect diversity relationship was generally stronger for low plant taxonomic ranks than at the level or orders or subclasses (Fig. 2). For most insect taxonomic ranks, and across most spatial scales, the relationship was strongest when the independent variable was the number of plant genera, and also quite strong for plant species. Plant phylogenetic diversity also performed well as a predictor of insect diversity, but not better than the number of plant genera (Fig. 2). The plant accumulation curves only approached saturation for the higher taxonomic ranks. In the case of subclasses, there was a clear decrease in slope at the 20 m scale, as well as at the 100 m and/or 1 km scale, with thicket having the highest number of plant subclasses at fine scales, but fynbos at the full biome scale. The pattern for plant orders was similar, but the decrease in slope at the 100 m scale was less pronounced, and thicket had the highest number of plant orders across scales. In the case of plant genera, the tendency towards saturation was only very slight, and largely limited to the 20 m scale (there was nearly no tendency for saturation in grassland, which was the richest at larger spatial scales). Plant species showed a similar decrease in slope at the 20 m scale, but an increase at the 1 km scale, especially in fynbos and grassland, the two most speciose biomes. The pattern in the accumulation of plant phylogenetic diversity was most similar to the one observed for plant genera, but with an additional tendency towards saturation at the 100 m and 1 km scales, albeit only slight (as compared to the strong tendency in taxonomic ranks higher than genus). The species accumulation curves for insects were mostly straight, with notable decreases in slope only at the 20 m scale, and grassland being most diverse across scales, followed by thicket and fynbos (Fig. 3). The four biomes also differed in terms of how plant diversity related to the diversity of insects in different feeding guilds. In thicket there was a strong relationship between plant diversity and herbivore diversity; in fynbos this relationship was weaker, but there was a strong relationship between plant diversity and detritivore diversity. In grassland and Nama-karoo there was no significant relationship between plant phylogenetic diversity and insect diversity in any of the guilds (Fig. 4). 4. Discussion The differences between biomes in the strength (and even sign) of the plant–insect diversity relationship suggest that different factors are the main drivers of insect diversity in each case. In Albany
96
Sß. Prochesß et al. / Molecular Phylogenetics and Evolution 51 (2009) 94–99 80
10m scale
50
10m scale insect species
insect species
60
40 30 20 10 0
60 40 20 0
0
10
20
30
40
0
50
500
1000
plant species
2500
2000
2500
20m scale
40 30 20 10 0
60
30
0 0
10
20
30
40
50
0
500
1000
plant species
1500
plant PD
80
135
100m scale
70 60
100m scale insect species
insect species
2000
90
20m scale
50
insect species
insect species
60
50 40 30 20 10 0
90
45
0 0
10
20
30
40
50
60
70
0
500
1000
plant species
1500
2000
2500
3000
plant PD
180
120
1km scale
100
1km scale insect species
insect species
1500
plant PD
80 60 40 20 0
135
fynbos
90
Grassland
45
Nama-karoo Albany thicket
0 0
20
40
60
80
100
plant species
0
500
1000
1500
2000
2500
3000
3500
plant PD
Fig. 1. The relationship between plant diversity and insect diversity in South African vegetation. Data for sixty-four 10 10 m plots in four biomes, pooled for larger spatial scales. Plant diversity quantified as species number and phylogenetic diversity (PD).
thicket, a stronger relationship at finer spatial scales (see Fig. 1) and particularly strong for herbivores (Fig. 4) indicates a direct connection between plant diversity and insect diversity. Indeed, thicket is the closest structurally to lowland tropical rainforest, in which previous studies on this topic indicated a direct relationship mediated by host plant specificity in insect herbivores (Novotny´ et al., 2006). In fynbos, the strong positive relationship between plant diversity and insect diversity is probably only partly the result of host specificity, and maybe partly relating to the role that finely-dissected fynbos plants play as heterogeneous habitats (especially for small insects; see Morse et al., 1985). The increased plant phylogenetic diversity noted at large spatial scales in fynbos (Fig. 3) more likely has to do with the inclusion of thicket elements, or with fynbos patches with higher soil fertility—these also being the reason for higher insect diversity (again, at large spatial scales). Heterogeneity could also be responsible for diversity patterns in the Nama-karoo, but different aspects of heterogeneity could be relevant for plants and insects, which would explain the absence of a clear plant–insect diversity connection in this biome. In grass-
land, the negative relationship between plant diversity and insect diversity could have to do with insects being more directly influenced by other factors such as productivity, and with higher productivity areas having fewer plant species. Indeed, these differences offer support to the recent observation of Lewinsohn and Roslin (2008) that high regional insect diversity can be the result of (a) high plant diversity, (b), high numbers of insect herbivores feeding on each plant, (c) high host specialization in herbivorous insects, or (d) high insect beta diversity. Although the species accumulation curves for insects in the four biomes look similar (Fig. 3), the plant–insect diversity plots (Fig. 1) suggest that the factors underlying this diversity may be different (high plant diversity in thicket (with low insect beta diversity, cf. rainforest in Novotny´ et al., 2007), high number of insects per plant in grassland and fynbos, and high beta diversity in fynbos and Nama-karoo (cf. plants in fynbos, Cowling and Lombard, 2002)). Insect beta diversity is likely to be much higher if including flightless local endemics, many of which are soil dwelling, and which are underrepresented in this study due to the collection method.
10m scale insect species insect families insect orders 20m scale insect species insect families insect orders 100m scale insect species insect families insect orders 1km scale insect species insect families insect orders biome-wide scale insect species insect families insect orders
*** *** ***
*** *** ***
*** *** *
** **
** **
*** *** ***
*** *** ***
*** *** ***
*** *** ***
** ** *
** ** *
** ** **
** ** ***
** *** **
* **
**
**
*
* *
*
97
plant subclasses
plant orders
*** *** ***
*
variation explained
significance *** ** *
plant families
plant genera
plant species
diversity
plant phylogenetic
Sß. Prochesß et al. / Molecular Phylogenetics and Evolution 51 (2009) 94–99
p<0.001 0.001
>75% 50-75% 25-50% 10-25% <10%
Fig. 2. Summary illustrating variations in the relationship between plant diversity and insect diversity in South African vegetation across spatial scales and taxonomic ranks. Data as in Fig. 1; percentage variation explained using linear models of log [insect diversity] against log [plant diversity]. The decrease in significance at large spatial scales is the result of low test power due to limited replication.
At large spatial scales (1 km), most of the variation in insect diversity is explained by plant diversity not within, but between biomes (cf. Figs. 1 and 2). The inclusion of the Nama-karoo biome, with much poorer plant and insect assemblages, is essentially responsible for the high percentage of variation explained. Thus, one needs to distinguish between (i) the high coincidence of plant diversity and insect diversity at large spatial scales (biome-wide), which is largely due to both variables responding similarly to the same environmental factors (Hawkins and Porter, 2003), as well as certain groups of plants and insects potentially having similar diversification histories, and (ii) the high number of insects directly influenced by the high number of plants (or high plant phylogenetic diversity) at low spatial scales (e.g. 20 m). To some extent, these two processes can be visualized as distinct peaks in the percentage of variation explained, with lower values at the intermediate 100 m and 1 km scales (Fig. 2). While there was no clear tendency towards diversity curve saturation (except in higher plant taxa, such as orders and subclasses—which were poor predictors of insect diversity), the slope of the plant diversity accumulation curves is potentially relevant to predicting insect diversity. Indeed, the curves for plant genera and plant phylogenetic diversity were most similar to the curve for insect species (Fig. 3), and these same measures of plant diversity were the best predictors of insect diversity (Fig. 2). The increase in slope for plant species at the 1 km scale indicates large dissimilarities between the two 1 km transects (e.g. in fyn-
bos, between the Baviaanskloof and the Kogelberg), and relates to the large number of local endemics in the fynbos and grassland biomes (Fig. 3), often in plant groups with poor dispersal (Cowling and Holmes, 1992). A similar level of local endemism would not be expected in flying insects, such as most species included in this study, but may occur in flightless insects. Thus, at large spatial scales, the number of plant genera and plant phylogenetic diversity may be the best predictors of flying insect species diversity, but the number of plant species could be a better predictor for flightless insects. These findings are relevant to the field of biodiversity estimation—in particular the fact that plant genera are a good predictor of insect species diversity across spatial scales (and not only for insect herbivores; see Supplementary data). Plant genera are quite easy to identify, and the use of plant generic diversity as a surrogate for insect diversity at the regional scale appears promising (cf. e.g. Lamoreux et al., 2006, for a discussion on the conservation implications and applications of biodiversity concordance). The results presented here need to be confirmed using a combination of collection methods, especially considering that flightless insects are underrepresented in sweep collections. The inclusion of a measure of insect phylogenetic diversity would add a much deeper perspective, as molecular sampling of insect diversity becomes more comprehensive. Furthermore, intraspecific genetic variation in plants, not covered here, could account for additional insect diversity (Bangert et al., 2006).
98
Sß. Prochesß et al. / Molecular Phylogenetics and Evolution 51 (2009) 94–99 200
120
90
T
T 60
30
10
1000
0 0
2
4
K
2000
K
K 60
3000
F plant PD
plant genera
plant species
110
TF G
G
G F
160
4000
6
0 0
2
log scale (m)
4
0
6
log scale (m) 16
30
2
4
6
log scale (m) 300
F T
F G
K 10
plant subclasses
plant orders
20
G
250
G
12
TF insect species
T
K 8
200
150
K 100
4 50
0
0 0
2
4
6
0 0
log scale (m)
2
4
6
log scale (m)
0
2
4
6
log scale (m)
Fig. 3. Accumulation curves for plant taxa, plant phylogenetic diversity (PD) and insect species in four Cape biomes (F, fynbos; G, grassland; K, Nama-karoo; T, Albany thicket). The lines are based on averaged values for all samples at every spatial scale (replication as in Figs. 1 and 2).
insect group
herbivores
detritivores predators/parasites
biome *
fynbos
**
grassland
Acknowledgments ***
Albany thicket Nama-karoo
variation explained
significance *** ** *
ingly often posted on the web (see e.g., Anon., 2006), conservationists may well be able to use these plant data to infer the diversity of the most diverse group of animals. When doing so, it will be important to partition insect diversity worldwide between rainforest and comparable systems, which have been so far the nearly single focus of global estimation approaches (e.g. Novotny´ et al., 2007), and regions with more limited insect alpha diversity, but possibly greater species turnover (see Lewinsohn and Roslin, 2008), such as the Cape Floristic Region.
p<0.001 0.001
>75% 50-75% 25-50% <10%
Fig. 4. Guild and biome differences in the relationship between plant phylogenetic diversity and insect species diversity at the 20 m scale (where the relationships was strongest overall; n = 8).
Broader comparisons using similar insect collection methods are needed, but ultimately, as botanical surveys become increas-
We thank the South African National Research Foundation, Stellenbosch University, the UK Darwin Initiative for the Survival of Species, the European Commission, the Royal Society (UK) and the Claude Leon Foundation for funding. Brigitte Braschler, Allan Ellis, Jan Schnitzler and John Wilson are thanked for discussion, and two anonymous reviewers for insightful comments. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ympev.2008.05.040.
References Anon., 2006. Acoraceae to Zosteraceae online. Nature 444, 136. APG, 2003. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG II. Bot. J. Linn. Soc. 141, 399–436.
Sß. Prochesß et al. / Molecular Phylogenetics and Evolution 51 (2009) 94–99 Bangert, R.K., Allan, G.J., Turek, R.J., Wimp, G.M., Meneses, N., Martinsen, G.D., Keim, P., Whitham, T.G., 2006. From genes to geography: a genetic similarity rule for arthropod community structure at multiple geographic scales. Mol. Ecol. 15, 4215–4228. Cowling, R.M., Holmes, P.M., 1992. Endemism and speciation in a lowland flora from the Cape Floristic Region. Biol. J. Linn. Soc. 47, 367–383. Cowling, R.M., Lombard, A.T., 2002. Heterogeneity, speciation/extinction history and climate: explaining regional plant diversity patterns in the Cape Floristic Region. Divers. Distrib. 8, 163–179. Finlay, B.J., Thomas, J.A., McGavin, G.C., Fenchel, T., Clarke, R.T., 2006. Self-similar patterns of nature: insect diversity at local to global scales. Proc. R. Soc. Lond. B 273, 1935–1941. Forest, F., Grenyer, R., Rouget, M., Cowling, R.M., Balmford, A., Davies, T.J., Faith, D.P., Manning, J.C., Prochesß, S ß ., van der Bank, M., Reeves, G., Hedderson, T.A.J., Savolainen, V., 2007. Preserving the evolutionary potential of floras in biodiversity hotspots. Nature 445, 757–760. García, L.V., 2004. Escaping the Bonferroni iron claw in ecological studies. Oikos 105, 657–663. Gering, J.C., Crist, T.O., 2002. The alpha-beta-regional relationship: providing new insights into local-regional patterns of species richness and scale dependence of diversity components. Ecol. Lett. 5, 433–444. Grimaldi, D., Engel, M.S., 2005. Evolution of The Insects. Cambridge University Press, New York. Hawkins, B.A., Porter, E.E., 2003. Does herbivore diversity depend on plant diversity? The case of California butterflies. Am. Nat. 161, 40–49. Johnson, S.D., 1992. Plant–animal relationships. In: Cowling, R.M. (Ed.), The Ecology of Fynbos: Nutrients, Fire and Diversity. Oxford University Press, Cape Town, pp. 175–205. Krüger, O., McGavin, G.C., 2001. Predator-prey ratio and guild constancy in a tropical insect community. J. Zool. 253, 265–273.
99
Lamoreux, J.F., Morrison, J.C., Ricketts, T.H., Olson, D.M., Dinerstein, E., McKnight, M.W., Shugart, H.H., 2006. Global tests of biodiversity concordance and the importance of endemism. Nature 440, 212–214. Lewinsohn, T.M., Roslin, T., 2008. Four ways towards tropical herbivore megadiversity. Ecol. Lett. 11, 398–416. Mittermeier, R.A., Robles-Gil, P., Hoffmann, M., Pilgrim, J.D., Brooks, T.M., Mittermeier, C.G., Lamoreux, J.L., Fonseca, G., 2004. Hotspots Revisited: Earth’s Biologically Richest and Most Threatened Terrestrial Ecoregions. Cemex & Conservation International, Mexico City. Morse, D.R., Lawton, J.H., Dodson, M.M., Williamson, M.H., 1985. Fractal dimension of vegetation and the distribution of arthropod body lengths. Nature 314, 731–733. Novotny´, V., Drozd, P., Miller, S.E., Kulfan, M., Janda, M., Basset, Y., Weiblen, G.D., 2006. Why are there so many species of herbivorous insects in tropical rainforests? Science 313, 1115–1118. Novotny´, V., Miller, S.E., Hulcr, J., Drew, R.A.I., Basset, Y., Janda, M., Setliff, G.P., Darrow, K., Stewart, A.J.A., Auga, J., Isua, B., Molem, K., Manumbor, M., Tamtiai, E., Mogia, M., Weiblen, G.D., 2007. Low beta diversity of herbivorous insects in tropical forests. Nature 448, 692–697. Prochesß, Sß., Cowling, R.M., 2006. Insect diversity in Cape fynbos and neighbouring South African vegetation. Global Ecol. Biogeogr. 15, 445–451. Prochesß, Sß., Wilson, J.R.U., Cowling, R.M., 2006. How much evolutionary history in a 10 10 m plot? Proc. R. Soc. Lond. B 273, 1143–1148. Stireman, J.O., Singer, M.S., 2003. What determines host range in parasitoids? An analysis of a tachinid parasitoid community. Oecologia 135, 629–638. Stork, N.E., 1987. Guild structure of arthropods from Bornean rain-forest trees. Ecol. Entomol. 12, 69–80. Wright, M.G., Samways, M.J., 1998. Insect species richness tracking plant species richness in a diverse flora: gall-insects in the Cape Floristic Region, South Africa. Oecologia 115, 427–433.
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev
Dissecting the plant–insect diversity relationship in the Cape Sß erban Prochesß a,b,c,*, Félix Forest d, Ruan Veldtman b,e, Steven L. Chown b, Richard M. Cowling a, Steven D. Johnson f, David M. Richardson b, Vincent Savolainen d,g a
Botany Department, Nelson Mandela Metropolitan University, P.O. Box 77000, Port Elizabeth 6031, South Africa Centre for Invasion Biology, Department of Botany and Zoology, Stellenbosch University, Private Bag X1, Matieland 7602, South Africa c School of Environmental Sciences, University of KwaZulu-Natal, Private Bag X54001, Durban 4000, South Africa d Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3DS, United Kingdom e South African National Biodiversity Institute, Kirstenbosch Research Centre, Private Bag X7, Claremont 7735, South Africa f School of Biological and Conservation Sciences, University of KwaZulu-Natal, Private Bag X01, Scottsville 3209, South Africa g Imperial College London, Silwood Park Campus, Buckhurst Road, Ascot, Berkshire, SL5 7PY, United Kingdom b
a r t i c l e
i n f o
Article history: Received 3 January 2008 Revised 23 May 2008 Accepted 27 May 2008 Available online 7 June 2008 Keywords: Accumulation curves Cape Floristic Region Diversity correlations Phylogenetic diversity Plant–insect interactions Spatial scale Taxonomic rank
a b s t r a c t It has been argued that insect diversity in the Cape is disproportionately low, considering the unusually high plant diversity in this region. Recent studies have shown that this is not the case, but the precise mechanisms linking plant diversity and insect diversity in the Cape are still poorly understood. Here we use a dated genus-level phylogenetic tree of the Cape plants to assess how plant phylogenetic diversity compares with taxonomic diversity at various levels in predicting insect diversity. We find that plant phylogenetic diversity (PD) is a better predictor of insect species diversity that plant species diversity, but the number of plant genera is overall as good a predictor as PD, and much easier to calculate. The relationship is strongest between biomes, suggesting that the relationship between plant diversity and insect diversity is to a large extent indirect, both variables being driven by the same abiotic factors and possibly by common diversification, immigration and extinction histories. However, a direct relationship between plant diversity and insect diversity can be detected at fine scales, at least within certain biomes. Diversity accumulation curves also indicate that the way plant phylogenetic diversity and the number of plant genera increase over spatial scales is most similar to that for insect species; plant species show a greater increase at large spatial scales due to high numbers of local endemics. Ó 2008 Elsevier Inc. All rights reserved.
1. Introduction An important step in the estimation of insect diversity worldwide was the recent finding that plant phylogenetic diversity is a good predictor of herbivorous insect species diversity (Novotny´ et al., 2006). However, this finding was based on phylogenetic trees including a very limited number of plant species. The comprehensive phylogenetic tree for Cape angiosperm genera (Forest et al., 2007) opens up new possibilities of analyzing the influence of plant phylogenetic diversity on insect diversity. In particular, the availability of a spatially explicit data set of plant and insect occurrences (Prochesß and Cowling, 2006) makes it possible to explore the influence of spatial scale on the relationship between insect diversity and plant diversity—while measuring the latter in terms of species, higher taxa, or cumulative phylogenetic distance. There are several aspects that can be clarified by combining these data. It is not clear whether the plant–insect diversity rela-
* Corresponding author. Address: School of Environmental Sciences, University of KwaZulu-Natal, Private Bag X54001, Durban 4000, South Africa. Fax: +27 31 2607317. E-mail address: [email protected] (S ß . Prochesß). 1055-7903/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2008.05.040
tionship holds for non-herbivorous insects and across different spatial scales, or whether plant phylogenetic diversity should be a predictor of any value for insects other than herbivores. Indeed, herbivores represent less than half of all insect species across a variety of ecosystems (Stork, 1987; Krüger and McGavin, 2001), and the diversity relationships between various insect feeding guilds are intricate (Stireman and Singer, 2003). The concordance between plant and insect diversity may hold across feeding guilds if plant and insect diversity respond similarly to environmental factors such as climate and topography (Hawkins and Porter, 2003). If this is the case, predictions of insect diversity patterns from plant diversity data would only be accurate across spatial scales that encompass substantial variation in climate or topography (Gering and Crist, 2002; Finlay et al., 2006). Here, we try to answer some of these questions by examining how the plant–insect diversity relationship varies with spatial scale, taxonomic rank and between insect feeding guilds. Our data come from one of the most plant–rich terrestrial areas of the world, i.e. four biomes (fynbos, grassland, Albany thicket and semi-arid Nama-karoo) across two global biodiversity hotspots in South Africa (the Cape Floristic Region and Maputaland–Pondoland–Albany; Mittermeier et al., 2004), although the magnitude of insect
Sß. Prochesß et al. / Molecular Phylogenetics and Evolution 51 (2009) 94–99
diversity in this region is disputed (Johnson, 1992; Wright and Samways, 1998; Prochesß and Cowling, 2006). 2. Methods The analyses are based on plant surveys and insect collections in 10- by 10-m plots. Insects were collected by exhaustively sweep-netting the vegetation. (Each plant or branch higher than 10 cm above ground was vigorously hit with the net at least once. All plants in the plots were sweep-netted, including shrubs and small trees; there were no large trees in any of the plots.) Eight plots were considered per locality, representing four co-linear sets of two adjacent plots, separated by 100 m, 1 km and 100 m. One locality for each biome was sampled in the Baviaanskloof Conservation Area in the Eastern Cape; additionally, for a second locality, fynbos plots were sampled in the Kogelberg Biosphere Reserve, Western Cape; grassland plots near Dordrecht in the Eastern Cape Drakensberg; Albany thicket plots in the Aloes Nature Reserve in Port Elizabeth, Eastern Cape; and Nama-karoo near Beaufort West, Western Cape. Insect collections were preformed during the season of maximum insect activity at each site: October in the Baviaanskloof and Kogelberg, February in the Drakensberg; December in Port Elizabeth; and April near Beaufort West. This sampling programme gave five spatial scales for analysis: 10 m (plot scale; n = 64); 20 m (pooled data for adjacent plots; n = 32); 100 m (pooled data for plots separated by distances of 100 m; n = 16); 1 km (pooled data for all plots in one locality; n = 8); and biome-wide (pooled data for the two localities in the same biome; n = 4). Plants were identified to species level, and insects assigned to morphospecies. In total, we listed 440 plant species and 636 insect morphospecies (59% phytophagous, 25% predaceous/parasitic and 16% detritivorous, on the basis of predominant feeding habits at family level). Both groups were assigned to higher ranks according to recent taxonomic treatments (APG, 2003; Grimaldi and Engel, 2005). In the case of plants, unplaced taxa that represent sister lineages to orders and subclasses were given order and subclass rank, respectively. Phylogenetic diversity (PD) was calculated from a tree including most angiosperm genera in the Cape (Forest et al., 2007). Sequences were not available for twenty-six of the 228 genera in our plots, and these were deleted from plant lists in calculating PD. This is unlikely to greatly affect the results, as most of them represented large families already present in every plot. The PD lost by not including intra-generic branch lengths is also likely to be minimal (Prochesß et al., 2006). At each spatial scale, each log-transformed insect diversity measure was regressed against each log-transformed plant diversity measure and the r2 fit of a linear model was extracted, as was the p value. The false detection rate method was used to correct for multiple comparisons (García, 2004). The results were found to be qualitatively consistent if using robust linear regressions or generalized linear models with Poisson errors. Accumulation curves for plant taxa, plant phylogenetic diversity, and insect species across spatial scales were drawn based on averaged values at each spatial scale. All analyses were implemented in the package R 2.3.1. (R Development Core Team; http://www.R-project.org). The robust linear regressions used function lmrob (library ‘robustbase’ version 0.2–3). 3. Results The plant–insect diversity relationship varied notably between biomes. In fynbos there was a consistent positive relationship between plant diversity and insect diversity, plant phylogenetic diversity being a better predictor of insect species diversity (as compared to the number of plant species). In grassland, the relationship be-
95
tween plant diversity and insect diversity was actually a negative one, at least at fine spatial scales. In Albany thicket, the relationship was positive, and plant species diversity and plant phylogenetic diversity performed equally well in predicting insect species diversity at fine scales, but, at the 1 km scale, the locality richer in plants was poorer in insects. There was no clear relationship between plant diversity and insect diversity in Nama-karoo (cf. Fig. 1). However, when analyzing data from all biomes together, the insect-plant diversity relationship was consistently positive across spatial scales and taxonomic ranks. The strength of this relationship varied as follows. First, the amount of variation in insect richness explained by plant richness increased with spatial scale (Fig. 2). This increase was most notable at large spatial scales when the predictor was the number of plant species, the number of plant genera, or plant phylogenetic diversity. When plant families, orders and subclasses were considered, there was a major increase in variation explained from the 10 m scale to the 20 m scale, and again from the 1 km scale to the biome-wide scale, but a slight decrease from 20 m to 1 km. For most taxonomic ranks, over 75% of the variation was explained at the biome scale, but this high percentage was largely due to the inclusion of the Nama-karoo biome, where both plant and insect diversity are much lower than in any of the other three biomes. Second, the plant–insect diversity relationship was generally stronger for low plant taxonomic ranks than at the level or orders or subclasses (Fig. 2). For most insect taxonomic ranks, and across most spatial scales, the relationship was strongest when the independent variable was the number of plant genera, and also quite strong for plant species. Plant phylogenetic diversity also performed well as a predictor of insect diversity, but not better than the number of plant genera (Fig. 2). The plant accumulation curves only approached saturation for the higher taxonomic ranks. In the case of subclasses, there was a clear decrease in slope at the 20 m scale, as well as at the 100 m and/or 1 km scale, with thicket having the highest number of plant subclasses at fine scales, but fynbos at the full biome scale. The pattern for plant orders was similar, but the decrease in slope at the 100 m scale was less pronounced, and thicket had the highest number of plant orders across scales. In the case of plant genera, the tendency towards saturation was only very slight, and largely limited to the 20 m scale (there was nearly no tendency for saturation in grassland, which was the richest at larger spatial scales). Plant species showed a similar decrease in slope at the 20 m scale, but an increase at the 1 km scale, especially in fynbos and grassland, the two most speciose biomes. The pattern in the accumulation of plant phylogenetic diversity was most similar to the one observed for plant genera, but with an additional tendency towards saturation at the 100 m and 1 km scales, albeit only slight (as compared to the strong tendency in taxonomic ranks higher than genus). The species accumulation curves for insects were mostly straight, with notable decreases in slope only at the 20 m scale, and grassland being most diverse across scales, followed by thicket and fynbos (Fig. 3). The four biomes also differed in terms of how plant diversity related to the diversity of insects in different feeding guilds. In thicket there was a strong relationship between plant diversity and herbivore diversity; in fynbos this relationship was weaker, but there was a strong relationship between plant diversity and detritivore diversity. In grassland and Nama-karoo there was no significant relationship between plant phylogenetic diversity and insect diversity in any of the guilds (Fig. 4). 4. Discussion The differences between biomes in the strength (and even sign) of the plant–insect diversity relationship suggest that different factors are the main drivers of insect diversity in each case. In Albany
96
Sß. Prochesß et al. / Molecular Phylogenetics and Evolution 51 (2009) 94–99 80
10m scale
50
10m scale insect species
insect species
60
40 30 20 10 0
60 40 20 0
0
10
20
30
40
0
50
500
1000
plant species
2500
2000
2500
20m scale
40 30 20 10 0
60
30
0 0
10
20
30
40
50
0
500
1000
plant species
1500
plant PD
80
135
100m scale
70 60
100m scale insect species
insect species
2000
90
20m scale
50
insect species
insect species
60
50 40 30 20 10 0
90
45
0 0
10
20
30
40
50
60
70
0
500
1000
plant species
1500
2000
2500
3000
plant PD
180
120
1km scale
100
1km scale insect species
insect species
1500
plant PD
80 60 40 20 0
135
fynbos
90
Grassland
45
Nama-karoo Albany thicket
0 0
20
40
60
80
100
plant species
0
500
1000
1500
2000
2500
3000
3500
plant PD
Fig. 1. The relationship between plant diversity and insect diversity in South African vegetation. Data for sixty-four 10 10 m plots in four biomes, pooled for larger spatial scales. Plant diversity quantified as species number and phylogenetic diversity (PD).
thicket, a stronger relationship at finer spatial scales (see Fig. 1) and particularly strong for herbivores (Fig. 4) indicates a direct connection between plant diversity and insect diversity. Indeed, thicket is the closest structurally to lowland tropical rainforest, in which previous studies on this topic indicated a direct relationship mediated by host plant specificity in insect herbivores (Novotny´ et al., 2006). In fynbos, the strong positive relationship between plant diversity and insect diversity is probably only partly the result of host specificity, and maybe partly relating to the role that finely-dissected fynbos plants play as heterogeneous habitats (especially for small insects; see Morse et al., 1985). The increased plant phylogenetic diversity noted at large spatial scales in fynbos (Fig. 3) more likely has to do with the inclusion of thicket elements, or with fynbos patches with higher soil fertility—these also being the reason for higher insect diversity (again, at large spatial scales). Heterogeneity could also be responsible for diversity patterns in the Nama-karoo, but different aspects of heterogeneity could be relevant for plants and insects, which would explain the absence of a clear plant–insect diversity connection in this biome. In grass-
land, the negative relationship between plant diversity and insect diversity could have to do with insects being more directly influenced by other factors such as productivity, and with higher productivity areas having fewer plant species. Indeed, these differences offer support to the recent observation of Lewinsohn and Roslin (2008) that high regional insect diversity can be the result of (a) high plant diversity, (b), high numbers of insect herbivores feeding on each plant, (c) high host specialization in herbivorous insects, or (d) high insect beta diversity. Although the species accumulation curves for insects in the four biomes look similar (Fig. 3), the plant–insect diversity plots (Fig. 1) suggest that the factors underlying this diversity may be different (high plant diversity in thicket (with low insect beta diversity, cf. rainforest in Novotny´ et al., 2007), high number of insects per plant in grassland and fynbos, and high beta diversity in fynbos and Nama-karoo (cf. plants in fynbos, Cowling and Lombard, 2002)). Insect beta diversity is likely to be much higher if including flightless local endemics, many of which are soil dwelling, and which are underrepresented in this study due to the collection method.
10m scale insect species insect families insect orders 20m scale insect species insect families insect orders 100m scale insect species insect families insect orders 1km scale insect species insect families insect orders biome-wide scale insect species insect families insect orders
*** *** ***
*** *** ***
*** *** *
** **
** **
*** *** ***
*** *** ***
*** *** ***
*** *** ***
** ** *
** ** *
** ** **
** ** ***
** *** **
* **
**
**
*
* *
*
97
plant subclasses
plant orders
*** *** ***
*
variation explained
significance *** ** *
plant families
plant genera
plant species
diversity
plant phylogenetic
Sß. Prochesß et al. / Molecular Phylogenetics and Evolution 51 (2009) 94–99
p<0.001 0.001
>75% 50-75% 25-50% 10-25% <10%
Fig. 2. Summary illustrating variations in the relationship between plant diversity and insect diversity in South African vegetation across spatial scales and taxonomic ranks. Data as in Fig. 1; percentage variation explained using linear models of log [insect diversity] against log [plant diversity]. The decrease in significance at large spatial scales is the result of low test power due to limited replication.
At large spatial scales (1 km), most of the variation in insect diversity is explained by plant diversity not within, but between biomes (cf. Figs. 1 and 2). The inclusion of the Nama-karoo biome, with much poorer plant and insect assemblages, is essentially responsible for the high percentage of variation explained. Thus, one needs to distinguish between (i) the high coincidence of plant diversity and insect diversity at large spatial scales (biome-wide), which is largely due to both variables responding similarly to the same environmental factors (Hawkins and Porter, 2003), as well as certain groups of plants and insects potentially having similar diversification histories, and (ii) the high number of insects directly influenced by the high number of plants (or high plant phylogenetic diversity) at low spatial scales (e.g. 20 m). To some extent, these two processes can be visualized as distinct peaks in the percentage of variation explained, with lower values at the intermediate 100 m and 1 km scales (Fig. 2). While there was no clear tendency towards diversity curve saturation (except in higher plant taxa, such as orders and subclasses—which were poor predictors of insect diversity), the slope of the plant diversity accumulation curves is potentially relevant to predicting insect diversity. Indeed, the curves for plant genera and plant phylogenetic diversity were most similar to the curve for insect species (Fig. 3), and these same measures of plant diversity were the best predictors of insect diversity (Fig. 2). The increase in slope for plant species at the 1 km scale indicates large dissimilarities between the two 1 km transects (e.g. in fyn-
bos, between the Baviaanskloof and the Kogelberg), and relates to the large number of local endemics in the fynbos and grassland biomes (Fig. 3), often in plant groups with poor dispersal (Cowling and Holmes, 1992). A similar level of local endemism would not be expected in flying insects, such as most species included in this study, but may occur in flightless insects. Thus, at large spatial scales, the number of plant genera and plant phylogenetic diversity may be the best predictors of flying insect species diversity, but the number of plant species could be a better predictor for flightless insects. These findings are relevant to the field of biodiversity estimation—in particular the fact that plant genera are a good predictor of insect species diversity across spatial scales (and not only for insect herbivores; see Supplementary data). Plant genera are quite easy to identify, and the use of plant generic diversity as a surrogate for insect diversity at the regional scale appears promising (cf. e.g. Lamoreux et al., 2006, for a discussion on the conservation implications and applications of biodiversity concordance). The results presented here need to be confirmed using a combination of collection methods, especially considering that flightless insects are underrepresented in sweep collections. The inclusion of a measure of insect phylogenetic diversity would add a much deeper perspective, as molecular sampling of insect diversity becomes more comprehensive. Furthermore, intraspecific genetic variation in plants, not covered here, could account for additional insect diversity (Bangert et al., 2006).
98
Sß. Prochesß et al. / Molecular Phylogenetics and Evolution 51 (2009) 94–99 200
120
90
T
T 60
30
10
1000
0 0
2
4
K
2000
K
K 60
3000
F plant PD
plant genera
plant species
110
TF G
G
G F
160
4000
6
0 0
2
log scale (m)
4
0
6
log scale (m) 16
30
2
4
6
log scale (m) 300
F T
F G
K 10
plant subclasses
plant orders
20
G
250
G
12
TF insect species
T
K 8
200
150
K 100
4 50
0
0 0
2
4
6
0 0
log scale (m)
2
4
6
log scale (m)
0
2
4
6
log scale (m)
Fig. 3. Accumulation curves for plant taxa, plant phylogenetic diversity (PD) and insect species in four Cape biomes (F, fynbos; G, grassland; K, Nama-karoo; T, Albany thicket). The lines are based on averaged values for all samples at every spatial scale (replication as in Figs. 1 and 2).
insect group
herbivores
detritivores predators/parasites
biome *
fynbos
**
grassland
Acknowledgments ***
Albany thicket Nama-karoo
variation explained
significance *** ** *
ingly often posted on the web (see e.g., Anon., 2006), conservationists may well be able to use these plant data to infer the diversity of the most diverse group of animals. When doing so, it will be important to partition insect diversity worldwide between rainforest and comparable systems, which have been so far the nearly single focus of global estimation approaches (e.g. Novotny´ et al., 2007), and regions with more limited insect alpha diversity, but possibly greater species turnover (see Lewinsohn and Roslin, 2008), such as the Cape Floristic Region.
p<0.001 0.001
>75% 50-75% 25-50% <10%
Fig. 4. Guild and biome differences in the relationship between plant phylogenetic diversity and insect species diversity at the 20 m scale (where the relationships was strongest overall; n = 8).
Broader comparisons using similar insect collection methods are needed, but ultimately, as botanical surveys become increas-
We thank the South African National Research Foundation, Stellenbosch University, the UK Darwin Initiative for the Survival of Species, the European Commission, the Royal Society (UK) and the Claude Leon Foundation for funding. Brigitte Braschler, Allan Ellis, Jan Schnitzler and John Wilson are thanked for discussion, and two anonymous reviewers for insightful comments. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ympev.2008.05.040.
References Anon., 2006. Acoraceae to Zosteraceae online. Nature 444, 136. APG, 2003. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG II. Bot. J. Linn. Soc. 141, 399–436.
Sß. Prochesß et al. / Molecular Phylogenetics and Evolution 51 (2009) 94–99 Bangert, R.K., Allan, G.J., Turek, R.J., Wimp, G.M., Meneses, N., Martinsen, G.D., Keim, P., Whitham, T.G., 2006. From genes to geography: a genetic similarity rule for arthropod community structure at multiple geographic scales. Mol. Ecol. 15, 4215–4228. Cowling, R.M., Holmes, P.M., 1992. Endemism and speciation in a lowland flora from the Cape Floristic Region. Biol. J. Linn. Soc. 47, 367–383. Cowling, R.M., Lombard, A.T., 2002. Heterogeneity, speciation/extinction history and climate: explaining regional plant diversity patterns in the Cape Floristic Region. Divers. Distrib. 8, 163–179. Finlay, B.J., Thomas, J.A., McGavin, G.C., Fenchel, T., Clarke, R.T., 2006. Self-similar patterns of nature: insect diversity at local to global scales. Proc. R. Soc. Lond. B 273, 1935–1941. Forest, F., Grenyer, R., Rouget, M., Cowling, R.M., Balmford, A., Davies, T.J., Faith, D.P., Manning, J.C., Prochesß, S ß ., van der Bank, M., Reeves, G., Hedderson, T.A.J., Savolainen, V., 2007. Preserving the evolutionary potential of floras in biodiversity hotspots. Nature 445, 757–760. García, L.V., 2004. Escaping the Bonferroni iron claw in ecological studies. Oikos 105, 657–663. Gering, J.C., Crist, T.O., 2002. The alpha-beta-regional relationship: providing new insights into local-regional patterns of species richness and scale dependence of diversity components. Ecol. Lett. 5, 433–444. Grimaldi, D., Engel, M.S., 2005. Evolution of The Insects. Cambridge University Press, New York. Hawkins, B.A., Porter, E.E., 2003. Does herbivore diversity depend on plant diversity? The case of California butterflies. Am. Nat. 161, 40–49. Johnson, S.D., 1992. Plant–animal relationships. In: Cowling, R.M. (Ed.), The Ecology of Fynbos: Nutrients, Fire and Diversity. Oxford University Press, Cape Town, pp. 175–205. Krüger, O., McGavin, G.C., 2001. Predator-prey ratio and guild constancy in a tropical insect community. J. Zool. 253, 265–273.
99
Lamoreux, J.F., Morrison, J.C., Ricketts, T.H., Olson, D.M., Dinerstein, E., McKnight, M.W., Shugart, H.H., 2006. Global tests of biodiversity concordance and the importance of endemism. Nature 440, 212–214. Lewinsohn, T.M., Roslin, T., 2008. Four ways towards tropical herbivore megadiversity. Ecol. Lett. 11, 398–416. Mittermeier, R.A., Robles-Gil, P., Hoffmann, M., Pilgrim, J.D., Brooks, T.M., Mittermeier, C.G., Lamoreux, J.L., Fonseca, G., 2004. Hotspots Revisited: Earth’s Biologically Richest and Most Threatened Terrestrial Ecoregions. Cemex & Conservation International, Mexico City. Morse, D.R., Lawton, J.H., Dodson, M.M., Williamson, M.H., 1985. Fractal dimension of vegetation and the distribution of arthropod body lengths. Nature 314, 731–733. Novotny´, V., Drozd, P., Miller, S.E., Kulfan, M., Janda, M., Basset, Y., Weiblen, G.D., 2006. Why are there so many species of herbivorous insects in tropical rainforests? Science 313, 1115–1118. Novotny´, V., Miller, S.E., Hulcr, J., Drew, R.A.I., Basset, Y., Janda, M., Setliff, G.P., Darrow, K., Stewart, A.J.A., Auga, J., Isua, B., Molem, K., Manumbor, M., Tamtiai, E., Mogia, M., Weiblen, G.D., 2007. Low beta diversity of herbivorous insects in tropical forests. Nature 448, 692–697. Prochesß, Sß., Cowling, R.M., 2006. Insect diversity in Cape fynbos and neighbouring South African vegetation. Global Ecol. Biogeogr. 15, 445–451. Prochesß, Sß., Wilson, J.R.U., Cowling, R.M., 2006. How much evolutionary history in a 10 10 m plot? Proc. R. Soc. Lond. B 273, 1143–1148. Stireman, J.O., Singer, M.S., 2003. What determines host range in parasitoids? An analysis of a tachinid parasitoid community. Oecologia 135, 629–638. Stork, N.E., 1987. Guild structure of arthropods from Bornean rain-forest trees. Ecol. Entomol. 12, 69–80. Wright, M.G., Samways, M.J., 1998. Insect species richness tracking plant species richness in a diverse flora: gall-insects in the Cape Floristic Region, South Africa. Oecologia 115, 427–433.