Content uploaded by Jerome K Vanclay
Author content
All content in this area was uploaded by Jerome K Vanclay
Content may be subject to copyright.
FBMIS Volume 1, 2004, 105-13 ISSN 1740-5955
http://www.fbmis.info/A/4_2_VanclayJ_1 Copyright @ 2003 The FBMIS Group
105
INDICATOR GROUPS AND FAUNAL RICHNESS
JEROME K VANCLAY
School of Environmental Science and Management, Southern Cross University
PO Box 157, Lismore NSW 2480, Australia [JVanclay@scu.edu.au]
(Submitted, 25th June 2003; Accepted, 19th May 2004; Published, 3rd June 2004)
ABSTRACT. Species richness is a popular indicator of ecosystem vitality, but is difficult
to assess. Many natural resource managers seek an efficient bioindicator, but the link
between candidate indicators and the richness of other taxononic groups remains elusive. A
series of faunal surveys in the Mbalmayo Forest Reserve in Cameroon suggest that it may
be possible to devise faunal bioindicators. The species richness of birds, of butterflies and of
termites is significantly correlated with total faunal richness across eight species groups,
suggesting that these groups may have potential as bioindicators, alone or in combination.
Although expensive, further research is warranted because of the substantial potential
benefits and implications of the use of indicator groups.
Keywords: alpha diversity, species richness, bioindicator, surrogate, butterflies, termites.
1 INTRODUCTION
Natural resource managers need a “canary” to draw attention to sites of special significance and to
forewarn them of impending problems (cf. the coal miner’s canary to warn of fatal methane levels). It
is impractical to comprehensively monitor every aspect of a resource; efficiency demands the use of
indicators as proxies (or surrogates in the sense of Prendergast et al. 1993) for comprehensive
assessment. The choice of indicator is critical, not only because of the inferences that may be drawn
from it, but also because an efficient indicator may free funds from monitoring for more productive
research, maintenance of the resource, and education of its constituency.
Researchers have considered many potential indicators (Brown 1991) or surrogates (Oliver and
Beattie 1996), including plant genera (e.g., Prance 1994), vegetative morphology (e.g., Gillison et al.
1996), vegetative structure (e.g., Ferris-Kaan et al. 1998), sound patterns (e.g., Riede 1993), birds
(Garson et al. 2003), insects (e.g., Halffter and Favila 1993, Kremen 1994) and rare species (Lawler et
al. 2003). While morphology and structure-based assessments may eventually offer reliable and
automated monitoring, many researchers resort to faunal indicators, assuming that their inter-
relationships with other fauna and flora will also extend to species richness. The expectation is that
species richness within a particular (often conspicuous) group should be correlated with the overall
faunal richness (and presumably also with vegetative richness), and thus that the welfare of the selected
indicator group should offer an insight into the state of the system as a whole. Unfortunately, there is
little empirical evidence to support the contention (e.g., Lindenmayer 1999, Ricketts et al. 2002,
Vessby et al. 2002).
Lawton et al. (1998) examined species richness (or alpha diversity as defined by Whittaker 1977) in
several animal groups (birds, butterflies, beetles, ants, termites, nematodes) sampled in the Mbalmayo
Forest Reserve, Cameroon (11°E, 3°N, 650 m above sea level) during 1992-94, and suggested that
assessments of habitat change based on familiar groups (e.g., birds, butterflies) may mislead because of
low pair-wise correlations between groups and weak trends with disturbance. Their conclusion may be
unnecessarily pessimistic, because such indicators may not be used to infer the richness within other
groups, but rather to gain an insight into overall species richness. An alternative interpretation of their
data offers a more promising prognosis, and does not exclude the possibility that some species groups
may indeed indicate overall faunal richness.
Lawton et al. (1998) found that 40 of 45 between-group correlations did not differ significantly
from zero (i.e., P>0.05), and thus found no reason to reject the null hypothesis that the species richness
within any one group bore no relationship to the richness in any other group. This is not the question
FBMIS Volume 1, 2004, 105-13 ISSN 1740-5955
http://www.fbmis.info/A/4_2_VanclayJ_1 Copyright @ 2003 The FBMIS Group
106
usually posed by resource managers, who often want to make inferences about the total species
richness. This question is explored below, using the Mbalmayo data kindly provided by Prof. John
Lawton.
2 DATA
The Mbalmayo data involve counts of individual species or morphospecies within nine taxonomic
groups sampled at six sites (Table 1). These observations have been adjusted to reflect an equal
sampling effort (Lawton et al. 1998), so some of the reported counts are fractional. Forty-five of the 54
possible site-species combinations were sampled, in most cases with a single sample, although two
samples were available in 15 instances. Nine site-species combinations remained unsampled. There are
two problems with these missing data: the column representing partial mechanical clearance, where
only 5 of the 9 species groups are sampled; and the row representing canopy ants, which were sampled
at only half of the sites. Despite this weakness, there are few better data presently available to address
this important and urgent question.
Table 1. Species counts from Mbalmayo Forest Reserve, Cameroon (Lawton et al. 1998).
Site and treatment †
Taxonomic group NP OS PCman PCmech CC FF
Birds 45 45 29 5 9
Butterflies 29, 33 51 30 28 30, 31 14
Malaise beetles ‡ 27, 31.5 40.5 33.5 32, 36.5 48.7
Intercept beetles ‡ 24, 47 113.5 41 59, 70.5 42.7
Canopy beetles 72 78 53, 80 91, 61 49, 46
Leaf litter ants 62, 55.3 73.6 79 72.8 46, 58.6
Termites 46 53 53 16 24
Soil nematodes 70.11 62.8 69.41 57.4 62.8, 67.36 54.05
Canopy ants 38.1, 28.8 35.7, 28.9 23.7, 31.8
† NP = Near-Primary forest, OS = Old-growth Secondary forest, PCman = Partly Cleared (manually) with
Terminalia ivorensis plantation 10-15 m tall, PCmech = Partly Cleared (mechanically) with T. ivorensis 10-15 m
tall, CC = Completely Cleared and planted to T. ivorensis 1-2 m tall, FF = manually cleared Farm Fallow (Lawson
et al. 1998). ‡ Malaise beetles = flying beetles caught in malaise traps; Intercept beetles = beetles caught in flight-
interception traps.
There are three ways to deal with the incomplete column (Table 1) representing partial mechanical
clearance: to omit the entire column, to pool it with the column representing partial manual clearance,
or to try to infer the missing values. The first option (omit) discards scarce information, and the third
option (infer missing values) involves making difficult and uncertain inferences, so the second option
was adopted. The two treatments involving partial clearance by manual and mechanical means are
similar in nature and in species counts (paired t-test, t6=0.6, P=0.6), and were combined. The row
representing canopy ants was omitted from the calculation of species totals (Table 2), but was included
in the analysis of possible species indicators.
Table 2. Maximum number of species recorded within each group at each site.
Taxon NP OS PC CC FF
Birds 45 45 29 5 9
Butterflies 33 51 30 31 14
Malaise beetles 31.5 40.5 33.5 36.5 48.7
Intercept beetles 47 113.5 41 70.5 42.7
Canopy beetles 72 78 91 49 0 †
Litter ants 62 73.6 79 58.6 60.5‡
Termites 46 53 53 16 24
Nematodes 70.1 62.8 69.4 67.4 54.1
Species total * 406.6 517.4 425.9 334 253
† Assumed to be zero, since no canopy. ‡ Interpolated from termite counts.
* Excludes canopy ants.
FBMIS Volume 1, 2004, 105-13 ISSN 1740-5955
http://www.fbmis.info/A/4_2_VanclayJ_1 Copyright @ 2003 The FBMIS Group
107
Multiple samples occur in 14 instances (after combining the two rows concerning partial clearing),
so there was more than one possible way to compute the total number of species for a site. For instance,
the total number of species at a site could be based on the average or the greatest of these multiple
observations of a species group at the site. In most cases, the difference between alternative
calculations was small, with the greatest discrepancy being canopy beetles on partially cleared sites
where the average (of 4 counts) was 71 and the maximum was 91 species. The present study based the
estimates of species totals on the greatest number of species observed, with the assumption that smaller
numbers were incomplete counts, and that the largest observations did not include vagrants (or
‘tourists’ in the sense of Moran and Southwood 1982). The possibility that the same beetle species may
occur in the malaise, flight interception and canopy data was dismissed, as these different trapping
methods catch different components of the beetle fauna (Lawton et al. 1998).
Two cells in the Table 2, canopy beetles and litter ants, were not sampled on the farm fallow site
and some assumptions were required to complete the table. Table 2 follows Lawton et al. (1998) in
assuming that no canopy beetles would be detected in farm fallow since no canopy was present at this
site. The number of litter ants was estimated through regression. Because of their similar niche and
reasonable correlation (r3=0.83, P=0.07), the likely number of litter ant species was estimated using a
simple linear regression of litter ants on termites (nants=50.2+0.42ntermites). These two inferred values
were used only to estimate the total number of species present on each site, and were not used directly
in further regression analyses.
3 ANALYSIS
The relationship between within-group richness and total faunal richness was examined using
regression analyses and permutation tests. Regression analyses were used to seek a predictor set of
organisms such that changes in the biological status of the predictor set reflect similar changes in a
wider group of organisms (Kitching 1993). Evidence of such qualities may be inferred from the
relationship Ni=ni + e, where e is a random error, ni is the number of species within group i, Ni is the
“extra-group” richness, the number of species in other groups Ni=Σj≠i nj, such that total surveyed
species richness is N=Ni+ni for anyi. Using total richness N as the response variable would artificially
enhance the quality of the fit (e.g., since N = ni+Ni = ni +e' even when no relationship between ni and
Ni exists). For the 14 site-species combinations where multiple samples were available, the individual
samples were used in further analyses, providing a total of 60 data observations (including the
observations on canopy ants, and excluding the presumed values for canopy beetles, canopy ants and
leaf litter ants in the farm fallow, see Appendix).
Figure 1. Extra-group richness (Ni) versus within-group richness (ni) in Mbalmayo
Forest Reserve, Cameroon. The solid line is the relationship Ni=246+3.5ni and the
dotted line represents the mean of the unfilled symbols, 342.
100
200
300
400
500
0 40 80 120
Within-group richness
Extra-group richness .
Birds Butterflies
Malaise beetles Intercept beetles
Canopy beetles Litter ants
Termites Nematodes
Canopy ants
FBMIS Volume 1, 2004, 105-13 ISSN 1740-5955
http://www.fbmis.info/A/4_2_VanclayJ_1 Copyright @ 2003 The FBMIS Group
108
Preliminary graphical analyses of these data (Figure 1) reveal that
1. the five sample sites are evident in these data as five bands, declining with slope -1 as within-group
richness increases (i.e., the relationship extra-group richness Ni = site richness Nsite minus within-
group species count ni),
2. some species groups (birds, butterflies and termites, illustrated with filled symbols) exhibit a
correlation between extra-group and within-group species numbers, and that
3. for some species (e.g., beetles caught in flight interception traps), the number of extra- and within-
group species appears uncorrelated.
Figure 2. Four species groups exhibit both large intercept and steep slope. Bars
indicate one standard error and illustrate the significance of the estimated slope.
Butterflies
Birds
Termites
Canopy ants
0
2
4
6
0 100 200 300
Intercept
Slope
These trends were confirmed by preliminary statistical analyses, which revealed three taxonomic
groups of interest (Table 3 and Figure 2): the birds, butterflies and termites, each of which has
relatively steep slope (β1), large intercept (β0), high correlation with extra-group richness (r), and a low
probability (P) that this is due to chance. Canopy ants exhibit a trend similar to these three groups
(Figure 2), but were recorded only at two sites (partially and completely cleared), so estimates for this
group are not significant (Table 3).
Table 3. Correlation coefficients for each species group.
Taxon n Sites β1 β0 r P
P
ˆ
Birds 5 5 3.4 271 0.76 0.06 0.06
Butterflies 8 5 5.8 180 0.81 0.006 0.006
Malaise beetles 7 5 -5.3 535 -0.43 0.16 0.17
Intercept beetles
7 5 0.4 303 0.15 0.37 0.40
Canopy beetles 8 4 1.3 257 0.44 0.13 0.15
Litter ants 7 4 2.9 158 0.62 0.06 0.07
Termites 5 5 3.8 203 0.77 0.05 0.07
Nematodes 7 5 4.1 61 0.3 0.25 0.23
Canopy ants 6 2 4.6 251 0.51 0.14 0.15
P is the probability derived from the correlation coefficient;
P
ˆis the estimated probability derived from a permutation test.
Tests of this kind may indicate a significant result purely due to chance. If the observed correlations
are due to chance alone, the associated probabilities should be uniformly distributed on [0,1]. However,
in this case, the probabilities observed remain small, suggesting that the correlation between within-
and extra-group richness is real (Figure 3, where the slope of the observed probabilities is 0.29,
significantly different from 1.0, P<0.001).
FBMIS Volume 1, 2004, 105-13 ISSN 1740-5955
http://www.fbmis.info/A/4_2_VanclayJ_1 Copyright @ 2003 The FBMIS Group
109
Figure 3. Probabilities reported in Table 3 are not distributed uniformly on [0,1].
0
0.2
0.4
0.6
0.8
1
0 0.2 0.4 0.6 0.8 1
Cumulative frequency
P
A further way to confirm the robustness of these findings is to resample (Good 2000). The
permutation test (
P
ˆ) reported in Table 3 results from shuffling the within- and between-group richness
data (Appendix) 1000 times, and reporting the relative number of times that the observed correlation
could have arisen by chance. The estimates from this test correspond closely to the conventional
probability estimates obtained from the single-sided t-test (Cohen 1977). Results of these tests are
illustrated in Figure 4. The curves indicate the correlations (in decreasing order) observed in the
shuffled data; different curves arise because of different numbers of observations, numbers of sites, and
random sequences for each species. There are only six instances (out of 1000) in the shuffled data that
exhibit a correlation higher than that observed in the real data (0.81), so it is unlikely that this
correlation is due to chance. In contrast, there are 397 instances in the shuffled data that exhibit a
correlation higher than that observed for intercept beetles, illustrating that this correlation may simply
be a chance occurrence.
Figure 4. Results of permutation tests
-1
0
1
0 1000
Number of instances
Correlation
Birds
Butterflies
Malaise beetles
Intercept beetles
Canopy beetles
Litter ants
Termites
Nematodes
Canopy ants
FBMIS Volume 1, 2004, 105-13 ISSN 1740-5955
http://www.fbmis.info/A/4_2_VanclayJ_1 Copyright @ 2003 The FBMIS Group
110
Regression analyses confirmed that it was appropriate to divide the data into two categories
(evidence for two categories F2,56=5.78, P=0.005; no evidence for additional categories F14,42=0.71,
P=0.8). One category contained four species groups, the birds, butterflies, termites, and canopy ants,
each of which exhibited a relationship with a slope of about 3.5 (evidence for positive slope in pooled
data: r=0.8, t22=3.5, P=0.0009). A simple linear regression (i.e., Ni=β0+β1ni) appeared adequate: a Box-
Cox analysis (Box and Cox, 1964, 1982) revealed no need to transform the response variable, and an
analysis of variance using six categories offered no evidence of a curvilinear relationship (F4,20=0.18,
P=0.9). The canopy ants were recorded on only two sites (PC and CC), so contribute little to the trend,
but lead to smaller residuals when included within the bird-butterfly-termite group than with the
remaining groups. The remaining category with five species groups exhibited no detectable trend (no
evidence for non-zero slope t34=0.93, P=0.2; and no evidence for a curvilinear relationship F2,34=1.05,
P=0.4). The resulting parameter estimates are given in Table 4.
Table 4. Parameter estimates to predict extra-group species richness from
within-group richness, Ni=β0+β1ni. Standard errors are shown in parentheses. All
parameters significant at P<0.01.
Species group Observations Intercept (β0) Slope (β1)
Birds, butterflies, termites,
& canopy ants
24 246.4
(27.6)
3.53
(0.81)
Other species groups 36 341.9
(11.9)
_
Butterflies and termites in
conjunction (Fig. 2)
5 133.8
(21.2)
2.20
(0.29)
Ideally, an indicator should have a high intercept (large and positive, because the “canary” should
die before the miners) and a steep slope (rich information content). However, the present data offer no
empirical way to discriminate between birds, butterflies and termites, as specific estimates of slope and
intercept for these groups do not differ significantly.
Figure 5. Extra-group richness (numbers of bird, beetle, litter ant and nematode species)
versus the number of butterfly plus termite species. The line represents the linear regression
Ni=133.8+2.2nI, where ni is the sum of butterfly and termite richness.
100
200
300
400
20 40 60 80 100
Richness of butterflies and termites
Extra-group richness
A “shopping basket” of selected surrogate taxa may form a better predictor set than a single species
group (e.g., di Castri et al. 1992, Kremen 1994, Oliver and Beattie 1996). Whilst the present data set is
too small to adequately resolve this issue, there is some evidence to support this contention and to draw
attention to the need for further research. For instance, butterflies and termites in conjunction provide a
good estimate of extra-group species richness (Table 4 and Figure 5). These estimates are based on the
regression of extra-group richness (total species minus the number of butterfly and termite species)
versus the numbers of termites plus the average of the numbers of butterfly counts reported in Table 1.
In this instance, there is a strong probability that the slope differs from zero (t3=7.59, P=0.002), even
FBMIS Volume 1, 2004, 105-13 ISSN 1740-5955
http://www.fbmis.info/A/4_2_VanclayJ_1 Copyright @ 2003 The FBMIS Group
111
after allowing a Bonferroni adjustment (Neyman and Pearson 1928, Stewart-Oaten 1995) for the three
possible pairwise combinations of birds, butterflies and termites (P=0.007), or the 36 possibilities of
pairing any of the groups (P=0.08). A permutation test that shuffled the data 10,000 times indicated an
estimated probability of 0.003.
4 DISCUSSION
The results reported in Table 3 and 4 obviously depend upon several assumptions, e.g., those
involved in:
1. collapsing Table 1 (i.e., assuming partial clearance by manual and mechanical means are not
significantly different),
2. completing missing cells in Table 2 (i.e., assuming no canopy beetles where there is no canopy, and
predicting richness of litter ants from termite richness),
3. estimating total surveyed faunal richness (i.e., using maximum rather than the average richness in
cells with multiple samples, omitting canopy ants from the total, assuming no beetle species occurs
in both interception and malaise traps), and
4. assuming that the faunal richness across the eight groups surveyed is indicative of the total richness
of all fauna (including fauna not sampled in the Mbalmayo study).
Fortunately, the findings appear to be relatively robust and hold when the first three of these
assumptions are varied, at least for the three groups involving birds, butterflies, and termites. When the
data were processed in other ways, consistent results were obtained for birds, butterflies and termites,
but canopy ants seemed more closely aligned with the second category of organisms under some
assumptions.
6 CONCLUSION
These findings support the contention (Garson et al. 2002) that some species groups (e.g., birds,
butterflies, and termites, in the case of Mbalmayo) may be useful indicators of the overall species
number. It seems that an even better indication of overall faunal richness may be obtained by using
diverse groups in conjunction (e.g., butterflies plus termites).
This observation must be qualified since the findings of the present study depend on the validity of
four assumptions made during the analyses (see above), and do not take into account the nature of these
species (viz. exotic versus endemic). It seems possible to make inferences about total species richness,
but one should not assume that all faunal groups follow the response of the chosen group (cf. Lawton et
al. 1998). These findings are specific to disturbed forest near Mbalmayo Forest Reserve in Cameroon,
and it should not be assumed that they are generally applicable. It seems feasible that butterflies and
termites in combination may be good indicators for other humid forest environments, but it is possible
that other species groups may be more effective indicators in other regions (e.g., arid or temperate
regions).
Faunal richness may not be a good indicator of vegetative disturbance, as some disturbances (e.g.,
partial clearance) may actually increase species richness (Table 2, and Lawton et al. 1998), and
disturbance can be gauged more easily and reliably in other ways (Watt 1998).
Because complete faunal inventories are difficult, time-consuming and expensive (Lawton 1998,
Stork 1995), most natural resource managers cannot monitor the status of all species operationally.
Many managers and researchers yearn for practical indicators that can be monitored efficiently and
extrapolated reliably. Several surrogates have been suggested, but little empirical evidence has been
tendered in support of these nominations. The present study offers some empirical evidence to support
the notion that selected species groups may serve as indicators of a broader group of fauna, particularly
when used in conjunction. If so, managers and researchers may be better served by reliable,
comprehensive studies of selected groups, rather than superficial attempts to survey the whole fauna.
However, the issue warrants further research (notably comprehensive faunal surveys for a range of
sites) since the potential benefits and implications are considerable.
FBMIS Volume 1, 2004, 105-13 ISSN 1740-5955
http://www.fbmis.info/A/4_2_VanclayJ_1 Copyright @ 2003 The FBMIS Group
112
REFERENCES
Brown, K.S., 1991. Conservation of neotropical environments: insects as indicators. In: N.M. Collins &
J.A. Thomas (eds) The Conservation of Insects and their Habitats, Academic, London, p. 349-404.
Box, G.E.P. and Cox, D.R., 1964. An analysis of transformations. Journal of the Royal Statistical
Society B 26:211-52.
Box, G.E.P. and Cox, D.R., 1982. An analysis of transformations revisited. Journal of the American
Statistical Association 77:209-210.
Castri, F.di, Robertson Vernhes, J. and Younes, T., 1992. Inventorying and monitoring biodiversity.
Biology International 27:1-27.
Cohen J., 1977. Statistical Analysis for the Behavioral Sciences, Lawrence Erlbaum, New Jersey.
Ferris-Kaan, R., Peace, A.J. and Humphrey, J.W., 1998. Assessing structural diversity in managed
forests. In: P. Bachmann, M. Köhl and R. Päivinen (eds) Assessment of Biodiversity for Improved
Forest Planning, Kluwer, Dordrecht. p. 331-342.
Garson, J., Aggarwal, A. and Sarkar, S., 2002. Birds as surrogates for biodiversity: an analysis of a data
set from southern Quebec. Journal of Biosciences 27(4):347-360.
Gillison, A.N., Liswanti, N. and Arief Rachman, I., 1996. Rapid Ecological Assessment of Kerinci
Seblat National Park Buffer Zone. CIFOR Working Paper No. 14, Dec 1996.
Good, P.I., 2000. Permutation Tests: A practical guide to resampling methods for testing hypotheses,
2nd ed, Springer, NY, 344 p.
Halffter, G. and Favila, M.E., 1993. The Scarabaeinae (Insecta: Coleoptera) an animal group for
analysing, inventorying and monitoring biodiversity in tropical rainforest and modified landscapes.
Biology International 27:15-21.
Kitching, R.L., 1993. Towards rapid biodiversity assessment – lessons following studies of arthropods
of rainforest canopies. In: Rapid Biodiversity Assessment: Proceedings of the biodiversity
assessment workshop, Macquarie University, Sydney. p. 26-30.
Kremen, C., 1994. Biological inventory using target taxa: a case study of the butterflies of Madagascar.
Ecological Applications 4:407-422.
Lawler, J.J., White, D., Sifneos, J.C. and Master, L.L., 2003. Rare species and the use of indicator
groups for conservation planning. Conservation Biology 17(3):875-882.
Lawton, J.H., Bignell, D.E., Bolton, B., Bloemers, G.F., Eggleton, P., Hammond, P.M., Hodda, M.,
Holt, R.D., Larsen, T.B., Mawdsley, N.A., Stork, N.E., Srivastava, D.S. and Watt, A.D., 1998.
Biodiversity inventories, indicator taxa and effects of habitat modification in tropical forest. Nature
391:72-76.
Lindenmayer, D.B., 1999. Future directions for biodiversity conservation in managed forests: indicator
species, impact studies and monitoring programs. Forest Ecology and Management 115:277-287.
Moran, V.C. and Southwood, T.R.E., 1982. The guild composition of arthropod communities in trees.
Journal of Animal Ecology 51:289-306.
Neyman, J., and Pearson, E.S., 1928. On the use and interpretation of certain test criteria for purposes of
statistical inference. Biometrika 20A:175-240, 263-97.
Oliver, I. and Beattie, A.J., 1996. Designing a cost-effective invertebrate survey: a test of methods for
rapid assessment of biodiversity. Ecological Applications 6:594-607.
Prance, G.T., 1994. A comparison of the efficacy of higher taxa and species numbers in the assessment
of biodiversity in the neotropics. Phil. Trans. R. Soc. London B 345:889-99.
Prendergast, J.R., Quinn, R.M., Lawton, J.H., Eversham, B.C. and Gibbons, D.W., 1993. Rare species,
the coincidence of diversity hotspots and conservation strategies. Nature 365:335-337.
Ricketts, T.H., Daily, G.C. and Ehrlich, P.R., 2002. Does butterfly diversity predict moth diversity?
Testing a popular indicator taxon at local scales. Biological Conservation 103(3):361-370.
Riede, K., 1993. Monitoring biodiversity: analysis of Amazonian rainforest sounds. Ambio 22:546-548.
Stewart-Oaten, A.1995. Rules and judgments in statistics: three examples. Ecology 76:2001-2009.
Stork, N.E., 1995. Measuring and inventorying arthropod diversity in temperate and tropical forests. In:
T.J.B. Boyle and B. Boontawee (eds) Measuring and Monitoring Biodiversity in Tropical and
Temperate Forests, CIFOR, Bogor, Indonesia. p. 257-270.
Vessby, K., Soderstrom, B., Glimskar, A. and Svensson, B., 2002. Species-richness correlations of six
different taxa in Swedish seminatural grasslands. Conservation Biology 16(2):430-439.
Watt, A.D., 1998. Measuring disturbance in tropical forests: a critique of the use of species-abundance
models and indicator measures in general. Journal of Applied Ecology 35:467-469.
Whittaker, R.H., 1977. Evolution of species diversity in land communities. In: M.K. Hecht, W.C. Steere
and B. Wallace (eds) Evolutionary Biology, Volume 10, Plenum, N.Y. p.1-67.
FBMIS Volume 1, 2004, 105-13 ISSN 1740-5955
http://www.fbmis.info/A/4_2_VanclayJ_1 Copyright @ 2003 The FBMIS Group
113
APPENDIX 1. DATA USED IN ANALYSES
Datum Site Group Within-group richness Extra-group richness
1 NP Birds 45 361.61
2 OS Birds 45 472.4
3 PCman Birds 29 396.91
4 CC Birds 5 328.96
5 FF Birds 9 243.98
6 NP Butterflies 29 377.61
7 NP Butterflies 33 373.61
8 OS Butterflies 51 466.4
9 PCman Butterflies 30 395.91
10 PCmech Butterflies 28 397.91
11 CC Butterflies 30 303.96
12 CC Butterflies 31 302.96
13 FF Butterflies 14 238.98
14 NP Malaise beetles 27 379.61
15 NP Malaise beetles 31.5 375.11
16 OS Malaise beetles 40.5 476.9
17 PCman Malaise beetles 33.5 392.41
18 CC Malaise beetles 32 301.96
19 CC Malaise beetles 36.5 297.46
20 FF Malaise beetles 48.7 204.28
21 NP Intercept beetles 24 382.61
22 NP Intercept beetles 47 359.61
23 OS Intercept beetles 113.5 403.9
24 PCman Intercept beetles 41 384.91
25 CC Intercept beetles 59 274.96
26 CC Intercept beetles 70.5 263.46
27 FF Intercept beetles 42.7 210.28
28 NP Canopy beetles 72 334.61
29 OS Canopy beetles 78 439.4
30 PCman Canopy beetles 53 372.91
31 PCman Canopy beetles 80 345.91
32 PCmech Canopy beetles 91 334.91
33 PCmech Canopy beetles 61 364.91
34 CC Canopy beetles 49 284.96
35 CC Canopy beetles 46 287.96
36 NP Litter ants 62 344.61
37 NP Litter ants 55.3 351.31
38 OS Litter ants 73.6 443.8
39 PCman Litter ants 79 346.91
40 PCmech Litter ants 72.8 353.11
41 CC Litter ants 46 287.96
42 CC Litter ants 58.6 275.36
43 NP Termites 46 360.61
44 OS Termites 53 464.4
45 PCman Termites 53 372.91
46 CC Termites 16 317.96
47 FF Termites 24 228.98
48 NP Nematodes 70.11 336.5
49 OS Nematodes 62.8 454.6
50 PCman Nematodes 69.41 356.5
51 PCmech Nematodes 57.4 368.51
52 CC Nematodes 62.8 271.16
53 CC Nematodes 67.36 266.6
54 FF Nematodes 54.05 198.93
55 PCman Canopy ants 38.1 425.91
56 PCman Canopy ants 28.8 425.91
57 PCmech Canopy ants 35.7 425.91
58 PCmech Canopy ants 28.9 425.91
59 CC Canopy ants 23.7 333.96
60 CC Canopy ants 31.8 333.96