Northeastern Naturalist Vol. 24, No. 1
K.W. McCravy
2017
15
2017 NORTHEASTERN NATURALIST 24(1):15–24
An Analysis of Malaise-Trap Effectiveness in Assessing
Robber Fly (Diptera: Asilidae) Species Richness
Kenneth W. McCravy1,*
Abstract - Little quantitative information is available on the effectiveness of Malaise traps in
estimating insect species richness. I used the Chao1 nonparametric species richness estimator
to evaluate the effectiveness of Malaise traps in assessing species richness of robber flies
in west-central Illinois burned and unburned deciduous forest and tallgrass prairie habitats.
Two of 12 traps yielded asymptotic species richness values, with remaining traps requiring
fold increases ranging from 3.1 to 20.9 to reach asymptotic richness. Overall, observed species
richness was over 85% of estimated species richness, but a 3.9-fold sample-size increase
would have been required to reach asymptotic species richness. For individual habitats, estimated
fold increases in sample size ranging from 1.9 to 9.7 would have been needed to reach
asymptotic species richness. These results show that Malaise traps are effective in sampling
robber flies, but asymptotic species richness would be difficult to achieve. Lower target coverage
levels may be practical in terms of sampling effort and cost, and use of complementary
sampling methods could improve the completeness of inventories.
Introduction
Anthropogenic habitat destruction and climate change are important drivers of
the Holocene mass extinction, otherwise known as the sixth extinction (Kolbert
2014). To conserve biodiversity and mitigate losses, knowledge of the present
state of biological diversity is needed. Such knowledge is only as reliable as the
sampling methods upon which it is based. Information on the reliability and limitations
of sampling methodology is essential for informed decisions in biological
conservation. However, it is often unclear how reliable are estimates of species
richness, and if a survey has sampled an acceptable proportion of the species
richness present, particularly for highly motile animals such as flying insects and
many vertebrates.
Insects are the most species-rich class of metazoans, and are ecologically and
economically important organisms in virtually all terrestrial environments. Studies
of the ecology, diversity, and conservation of insects rely on efficient means of insect
sampling. A variety of methods for sampling flying insects have been developed,
including light traps, suction traps, pan traps, sticky traps, baited traps, interception
traps, and Malaise traps (Young 2005). Initially proposed by Malaise (1937), several
Malaise-trap designs have been developed (Ozanne 2005), the most popular of
which is probably the Townes-style Malaise trap (Townes 1972). The Townes-style
Malaise trap has been used to sample general insect diversity (e.g., Deans et al. 2005,
1Department of Biological Sciences, Western Illinois University, One University Circle,
Macomb, IL 61455. *Corresponding author - KW-McCravy@wiu.edu.
Manuscript Editor: Christopher M. Heckscher
Northeastern Naturalist
16
K.W. McCravy
2017 Vol. 24, No. 1
Richards and Windsor 2007) as well as specific insect taxa such as robber flies (e.g.,
McCravy and Baxa 2011), horse flies and deer flies (e.g., Roberts 1978), parasitic
wasps (e.g., Fraser et al. 2008, Skillen et al. 2000), sphecine wasps (e.g., McCravy et
al. 2009), bees (e.g., Geroff et al. 2014), butterflies (e.g., Covell and Freytag 1979),
beetles (e.g., Harris and Burns 2000), and ground beetles (e.g., Ulyshen et al. 2005).
However, while numerous studies have used Malaise traps to estimate insect species
richness, relatively few studies have addressed the reliability of these estimates. For
example, a few studies have assessed Malaise-trap effectiveness in measuring ichneumonid
wasp richness, and found that asymptotic species richness was generally
not achieved (Di Giovanni et al. 2015, Fraser et al. 2008, Skillen et al. 2000). Recent
studies by Geroff et al. (2014) and McCravy et al. (2016) suggest that Malaise traps
can be an effective means of assessing native bee species richness, but that the usefulness
of these traps for bee-diversity assessments may be limited by the cost of the
traps and the substantial variability in the results produced by individual traps. However,
such information is lacking for most insect taxa.
Robber flies (Diptera: Asilidae) are a diverse family of predatory insects, with
over 7000 described species worldwide (Ghahari et al. 2007) and ~1000 in the Nearctic
region (Poole 1996). A total of 80 species are listed for Illinois (Raney 2003),
of which 28 have been collected in the west-central Illinois counties of Hancock and
McDonough by Kartawich and McCravy (2010) and McCravy and Baxa (2011).
Many robber fly species have specific habitat requirements (McCravy and Baxa
2011), which, along with their role as top insect predators, makes robber flies a
taxon of special conservation concern (Barnes et al. 2007). Additionally, many
robber fly species are large in size and distinctive in appearance, making the family
potentially useful for conservation and biodiversity research (Larson and Meier
2004). In this paper, I investigate the effectiveness of Townes-style Malaise traps
in assessing the species richness of robber flies in a Midwestern restored tallgrass
prairie and deciduous forest, using a reanalysis of data collected by McCravy and
Baxa (2011).
Field-Site Description
This study was done in May to October 2005 at Alice L. Kibbe Life Science
Station in Hancock County, IL. The station includes ~90 ha owned by Western
Illinois University. An additional ~590 ha owned by the Illinois Department of
Natural Resources is located adjacent to the field station, and together these 2
holdings comprise the Kibbe Macro Site. The Kibbe Macro Site contains a variety
of vegetation communities including mature upland oak–hickory (Quercus
spp.–Carya spp.) forests, mature floodplain forests, early successional forests, oak
barrens, hill prairies, and restored tallgrass prairies (Western Illinois University
2016). Sampling for this study was conducted in 4 habitat types: restored tallgrass
prairie burned in spring 2005 (henceforth, referred to as “burned prairie” or “BP”;
40°22'0"N, 91°24'26"W), restored tallgrass prairie last burned the previous year, in
spring 2004 (“unburned prairie,” or “UP”; 40°21'49"N, 91°24'16"W), oak–hickory
forest burned in spring 2005 (“burned forest,” or “BF”; 40°21'59"N, 91°24'28"W),
Northeastern Naturalist Vol. 24, No. 1
K.W. McCravy
2017
17
and an oak–hickory forest unburned for >5 years (“unburned forest,” or “UF”;
40°21'46"N, 91°24'06"W). These sites have been previously described by McCravy
and Baxa (2011). The following summary is provided for convenience.
The prairie habitats had been agricultural fields until the late 1970s, when they
were restored. Each prairie habitat was ~3 ha. The burned prairie was ~500 m
in length and ~60 m wide, whereas the unburned prairie was irregular in shape.
Common plants in the prairie habitats included Sorghastrum nutans (L.) Nash
(Indian Grass), Andropogon gerardii Vitman (Big Bluestem), Schizachyrium
scoparium (Michx.) Nash (Little Bluestem), Panicum virgatum L. (Switchgrass),
and various forbs (Asteraceae and Fabaceae). Common woody species included
Cornus drummondii C.A. May (Rough-leaved Dogwood) and Rhus glabra L.
(Smooth Sumac). Mean percent ground cover (± SE) was measured by visual
estimation of twenty 0.5-m2 plots in each prairie habitat in mid-July and mid-
September 2005. In mid-July, mean percent ground cover was 55.5 ± 3.73 in the
burned prairie and 80.5 ± 3.80 in the unburned prairie, respectively. In mid-September,
mean percent ground cover was 71.0 ± 2.98 in the burned prairie and 76.5
± 2.54 in the unburned prairie, respectively.
The forest habitats were old-growth, dry-mesic upland oak–hickory forests that
were open woodland–savanna with a history of grazing until the 1950s–1960s.
After this time, fire suppression allowed the emergence of a dense understory. Beginning
in the mid-1990s, the burned forest underwent prescribed burning every
2–3 years in order to suppress understory vegetation. The unburned forest had a
history of more-infrequent and inconsistent fire and had not been burned in >5
years. Total contiguous forest area was at least 200 ha, with burn units of ~3 ha
each. Quercus alba L. (White Oak), Quercus rubra L. (Northern Red Oak) and
Carya ovata (Miller) K. Koch (Shagbark Oak) were dominant overstory tree species.
Mean percent ground cover (± SE) in the forest habitats was also measured
by visual estimation of twenty 0.5-m2 plots. In mid-July 2005, these values were
29.0 ± 3.24 and 59.5 ± 3.36 in the burned forest and unburned forest, respectively.
In mid-September 2005, these measures were 23.5 ± 2.84 and 38.0 ± 3.88, respectively.
Plant nomenclature follows that of Gleason and Cronquist (1991).
Methods
Data for this study were collected as part of an assessment of robber fly diversity
and habitat associations at Kibbe Life Science Station, and a list of species collected
has been previously published (McCravy and Baxa 2011). With help from a
student researcher, I placed 3 Townes-style Malaise traps (Sante Traps, Lexington,
KY) in each of 4 habitats (burned prairie, unburned prairie, burned forest, and
unburned forest), spacing them a minimum of 75 m apart and a maximum of 250
m apart within each habitat. Traps located in prairies were 15 m to 30 m from the
nearest forest/prairie interface, and traps located in forests were 25 m to 50 m from
the nearest forest/prairie interface. We filled trap collection bottles with 75% ethyl
alcohol and operated the traps continuously from 16 May to 24 October 2005. We
collected samples every 3–4 days. Robber flies were collected from the samples,
Northeastern Naturalist
18
K.W. McCravy
2017 Vol. 24, No. 1
pinned, labeled, and identified using reference specimens and the identification key
in Wood (1981). We deposited voucher specimens in the Western Illinois University
Entomology Collection.
Total numbers (n), observed species richness (Sobs), and estimated asymptotic
species richness (Sest) of robber flies were determined for each trap, all traps combined
within a habitat, and all traps combined, with collections summed over the
entire season. I calculated values of Sest using the Chao1 nonparametric species richness
estimator (Ecological Archives E090-073-S1, Chao et al. 2009). The Chao1
estimator produces an estimate of the minimum species richness present, based on
the relative numbers of singletons (1 individual of a species collected) and doubletons
(2 individuals of a species collected) (Chao et al. 2009). I also used the Chao1
estimator to evaluate the completeness with which Malaise traps sampled robber fly
species richness by calculating probabilities that an additional individual sampled
would represent a previously undetected species (the proportion of singletons in
the sample: q0 = f1 / n), and sample sizes needed to achieve 80%, 90%, 95%, and
100% of Chao1 estimates. For each Chao1 estimate, the proportion of singletons
was less than 50% (i.e., f1 / n < 0.5), as recommended by Anne Chao (cited in Colwell
2013). I conducted single-classification analysis of variance (ANOVA; ɑ =
0.05) using SigmaPlot 13.0 (Systat Software Inc., San Jose, CA) to compare mean
numbers of robber flies collected, mean Sobs, and mean Sest among the 4 habitats. For
each analysis, I evaluated assumptions of normality and equal variances using the
Shapiro-Wilk test and the Brown-Forsythe test, respectively. Assumptions were met
in all cases (mean number of robber flies: P = 0.476 and P = 0.303, respectively;
mean Sobs: P = 0.916 and P = 0.138, respectively; mean Sest: P = 0.122 and P = 0.684,
respectively). If significant differences were detected among means, I compared the
means using the Holm-Šidák multiple comparison procedure (ɑ = 0.05). Withinhabitat
coefficients of variation ([SD / mean] x 100) were calculated for Sobs and Sest.
Results
A total of 668 robber flies representing 26 species were collected overall. The
most common species, comprising 63.3% of the total, were Nerax aestuans (L.)
(20.8%), Promachus hinei Bromley (12.1%), Psilonyx annulatus (Say) (10.8%),
Ommatius ouachitensis Bullington & Lavigne (10.2%), and Atomosia glabrata
(Say) (9.4%). The complete list of species can be found in McCravy and Baxa
(2011). A total of 201 individuals and 17 species were collected in the burned forest,
115 individuals and 15 species in the unburned forest, 248 individuals and 17
species in the burned prairie, and 104 individuals and 14 species in the unburned
prairie (Table 1).
Numbers of robber flies collected by individual Malaise traps ranged from 24
for UF #1 to 89 for BP #1 (Table 1). Mean number of robber flies collected per trap
overall (± SE) was 55.67 ± 6.73, and ranged from 34.67 ± 3.48 in the unburned
prairie to 82.67 ± 5.36 in the burned prairie. There was a significant difference (F =
9.721; df = 3, 11; P = 0.005) in the mean numbers of robber flies collected among
habitats (Fig. 1). Traps in the burned prairie collected significantly more robber
Northeastern Naturalist Vol. 24, No. 1
K.W. McCravy
2017
19
Table 1. Abundance (n), observed (Sobs) and estimated (Sest) species richness, probability that an additional
individual sampled would be a previously undetected species (q0), and estimated sample sizes
needed to achieve a given percentage of Sest for robber flies (Asilidae) from 12 Malaise traps operated
from May to October 2005 in four habitats at Alice L. Kibbe Life Science Station, Hancock County,
IL. BF = burned forest, UF = unburned forest, BP = burned prairie, UP = unburned prairie.
Estimated # of robber flies (and fold increase)
required to achieve given % of Sest
Trap n Sobs Sest q0 80% 90% 95% 100%
BF #1 52 11 11.00 0.000 -A -A -A -A
BF #2 65 13 17.50 0.046 90 (1.4) 157 (2.4) 225 (3.5) 422 (6.5)
BF #3 84 14 18.00 0.048 93 (1.1) 151 (1.8) 209 (2.5) 387 (4.6)
BF Total 201 17 25.00 0.020 390 (1.9) 669 (3.3) 947 (4.7) 1950 (9.7)
UF #1 24 10 12.67 0.167 25 (1.0) 36 (1.5) 47 (2.0) 76 (3.2)
UF #2 38 11 35.50 0.184 203 (5.3) 295 (7.8) 387 (10.2) 793 (20.9)
UF #3 53 9 9.25 0.019 -A -A -A -A
UF Total 115 15 16.13 0.026 -B -B 129 (1.1) 220 (1.9)
BP #1 89 9 12.00 0.034 -C -C -C -C
BP #2 87 13 15.67 0.046 -D 118 (1.4) 158 (1.8) 274 (3.1)
BP #3 72 11 35.50 0.097 384 (5.3) 559 (7.8) 733 (10.2) 1503 (20.9)
BP Total 248 17 26.00 0.024 452 (1.8) 710 (2.9) 968 (3.9) 1937 (7.8)
UP #1 41 7 9.00 0.049 45 (1.1) 74 (1.8) 102 (2.5) 151 (3.7)
UP #2 34 12 16.17 0.147 41 (1.2) 61 (1.8) 80 (2.4) 139 (4.1)
UP #3 29 9 19.00 0.172 -C -C -C -C
UP Total 104 14 18.17 0.048 116 (1.1) 176 (1.7) 236 (2.3) 426 (4.1)
All Traps 668 26 30.50 0.009 -D 863 (1.3) 1210 (1.8) 2592 (3.9)
ASobs was 100% of Sest (rounded down to 9.00 for UF #3).
BSobs was greater than 90% of Sest.
CCould not be calculated because no doubletons were collected.
DSobs was greater than 80% of Sest.
Figure 1. Numbers of
robber flies (Asilidae)
(mean ± 1 SE) collected
by Malaise traps in 4
habitats at the Alice L.
Kibbe Life Science Station,
Hancock County,
IL, in May to October
2005. Means with
the same letter are not
significantly different
(Holm-Šidák multiple
comparison procedure,
ɑ = 0.05); n = 3 Malaise
traps per habitat.
Northeastern Naturalist
20
K.W. McCravy
2017 Vol. 24, No. 1
flies than those in the unburned forest (P = 0.015) and unburned prairie (P = 0.008).
Sobs ranged from 7 for UP #1 to 14 for BF #3. Mean Sobs per trap overall was 10.75
± 0.59, and ranged from 9.33 ± 1.45 in the unburned prairie to 12.67 ± 0.88 in the
burned forest. There was no significant difference in mean Sobs among habitats
(F = 1.846; df = 3, 11; P = 0.217). Coefficients of variation for Sobs were 12.1% for
burned forest traps, 10.0% for unburned forest traps, 18.2% for burned prairie traps,
and 27.0% for unburned prairie traps.
Overall Sest was 30.5 (Table 1). For individual traps, Sest ranged from 9.0 for UP
#1 to 35.5 for UF#2 and BP #3. Sest for specific habitats ranged from 16.13 for the
unburned forest to 26.0 for the burned prairie. There was no significant difference in
mean Sest among habitats (F = 0.267; df = 3, 11; P = 0.848). Coefficients of variation
for Sest were 25.2% for burned forest traps, 74.6% for unburned forest traps, 60.0%
for burned prairie traps, and 35.0% for unburned prairie traps.
The overall value of q0 was 0.009, and ranged from 0.000 for BF #1 to 0.184 for
UF #2 (Table 1). Among different habitats, q0 ranged from 0.020 for the burned forest
to 0.048 for the unburned prairie. To reach the overall estimated asymptotic Sest
of 30.50, an estimated sample size increase of 3.9-fold would be required (Table 1).
For individual traps, estimated sample size increases needed to reach asymptotic
Sest ranged up to 20.9-fold for UF #2 and BP #3 (Sest was reached for BF #1 and UF
#3). For different habitats, estimated required sample size increases to achieve Sest
ranged from 1.9-fold for the unburned forest to 9.7-fold for the burned forest.
Discussion
The Sobs for all traps combined (26) was more than 85% of the Sest of 30.50,
suggesting that Malaise traps were effective in collecting most of the robber fly
species likely to be sampled using this method. Proportions of singletons (q0) collected
by individual traps ranged from 0 to 0.184. These numbers were similar to
q0 values for bees collected in restored tallgrass prairie at Kibbe Field Station (Geroff
et al. 2014). In that study, q0 values for individual Malaise traps ranged from
0.012 to 0.150 (K. McCravy, unpubl. data). In a study of vertical stratification of
ichneumonid wasp communities for which individual Malaise trap (type unspecified)
collection numbers were given, q0 ranged from 0.007 to 0.017 for understory
traps, and 0.083 to 0.270 for canopy traps (q0 values calculated from data in Table 1
of Di Giovanni et al. 2015). The low q0 values for understory traps in that study may
have been a result of very low species evenness. A single species accounted for 71%
to 85% of collections for individual traps.
Substantial increases in trapping effort would be needed to approach asymptotic
robber fly species richness at Kibbe Life Science Station, as reflected by the
estimated overall required sample size of 2592, or 1924 additional robber flies, a
3.9-fold increase over the actual sample size. In part, this reflects the increasing
difficulty of collecting very rare species. Based on the relatively low q0 of 0.009,
an estimated 111 additional robber flies would be required to yield 1 previously
uncollected species, using the geometric distribution 1 / P. Sample size increases
of 1.9 to 9.7 would be required to reach asymptotic species richness in specific
Northeastern Naturalist Vol. 24, No. 1
K.W. McCravy
2017
21
habitats. These numbers were generally lower than those required for complete
coverage of bee species richness in restored tallgrass prairie at the same site in
the study of McCravy et al. (2016). These results probably reflect the much lower
richness of robber flies at the site. Fraser et al. (2008) found that even 16 Malaise
traps within a single wooded habitat were insufficient to completely sample 4
subfamilies of Ichneumonidae.
Within habitats, the overall Sobs was consistently higher than that of any individual
trap, indicating that multiple traps increased the likelihood of collecting
greater numbers of species. But results of the present study indicate that achieving
a complete inventory of robber fly species richness would probably be impractical
from a trapping-effort standpoint, particularly considering the potentially prohibitive
cost of deploying large numbers of Malaise traps. Lower thresholds, such as
80% or 90% of Sest, would be a more realistic goal. In one habitat, unburned forest,
over 90% of Sest was obtained. For the other habitats, estimated sample size fold
increases ranging from 1.7 to 3.3 would have been needed to reach 90% of Sest, and
a 1.3-fold increase would have been needed to reach 90% of overall Sest. Sest and q0
can be used to establish predetermined “stopping rules”, a point beyond which
further sampling would be unnecessary or too costly (Magurran 2004). However,
this approach would require ongoing identification of species during the sampling
period, which is often not practical when large numbers of difficult-to-identify
specimens are involved.
The Sest of individual malaise traps varied widely. Coefficients of variation for
Sest were substantially higher than those for Sobs for each of the 4 habitats. These
results show that Malaise trap Sest values can vary widely among traps, even among
those in the same habitat with relatively low variation in Sobs. Thus, even having
several Malaise traps produce relatively consistent Sobs values in a given habitat
may not be a reliable indicator of actual species richness or of the level of variation
of species richness.
As is the case with other trapping methods and taxa, behavioral peculiarities
of particular robber fly species may increase or decrease the relative likelihood of
capture in Malaise traps. Such biases have been shown for epigeal spiders in pitfalltrap
collections (Topping 1993) and for bees with a variety of trap types, including
Malaise traps (Geroff et al. 2014). It is therefore probable that some robber fly species
present at the study site are relatively unlikely to be collected by Malaise traps.
Combining Malaise traps with another collection method, such as active aerial netting,
would likely provide a more complete robber fly species inventory than use
of a single sampling method, although aerial netting can be labor-intensive and has
inherent biases as well. Geroff et al. (2014) found that multiple sampling methods
produced a more complete inventory of bee diversity than did any single method.
More research on differences in relative performance and complementarity of different
sampling methods, and the advantages of using multiple methods, is needed
for other arthropod taxa.
The results of this study suggest the possibility that Malaise trap performance
is habitat dependent. In forest and prairie habitats, the burned habitats produced
Northeastern Naturalist
22
K.W. McCravy
2017 Vol. 24, No. 1
greater numbers of robber flies, significantly so in the case of the burned prairie.
These results may reflect true differences in robber fly densities in these habitats.
But because results using Malaise traps are a function of insect density and activity
levels, these results could also reflect habitat-dependent variation in robber fly
activity. In particular, the modified vegetation structure of burned habitats may
mean that robber flies in those habitats spend a greater proportion of time in flight
rather than perched on vegetation, and/or have more open flight paths available,
making those flies more likely to be captured in Malaise traps. Robber fly species
composition varied among habitats at this location, and some robber fly species
were habitat-specific (McCravy and Baxa 2011), so Malaise trap performance may
be related to the behavior and activity patterns of particular species as well. Inclusion
of well-designed aerial net sampling could compensate for such biases.
Assessment of species richness is a fundamental goal of many conservation
efforts, and Malaise traps are frequently used in inventories of flying insect taxa.
However, there is little information on the effectiveness of Malaise traps in assessing
species richness. Results of this study suggest that Malaise traps offer an
effective method of sampling robber flies, and can provide an acceptable inventory,
depending on the target coverage level. As with any sampling method, Malaise
traps probably have species-specific biases, which could be addressed by incorporating
other sampling methods into the inventory. Cost is a consideration in using
Malaise traps, but because Malaise traps can collect large numbers of a wide variety
of taxa, it is also possible that the additive value of assessing multiple insect
taxa could make Malaise traps more economically advantageous. More research is
needed on the effectiveness of Malaise traps in measuring species richness of different
insect taxa under various environmental conditions.
Acknowledgments
I thank Krista Baxa (Aledo, IL) for assistance with robber fly collection, pinning, and
labeling, and Dr. Herschel Raney (Conway, AR) for assistance with robber fly identifications.
Dr. Seán Jenkins (Western Illinois University) provided helpful information on habitat
management and plant species present at the Alice L. Kibbe Life Science Station. Two
anonymous reviewers provided helpful suggestions and comments on the manuscript. The
Western Illinois University Centennial Honors College and College of Arts and Sciences
provided generous funding in support of this project.
Literature Cited
Barnes, J.K., N. Lavers, and H. Raney. 2007. Robber flies (Diptera: Asilidae) of Arkansas,
USA: Notes and a checklist. Entomological News 118:241–258.
Chao, A., R.K. Colwell, C-W. Lin, and N.J. Gotelli. 2009. Sufficient sampling for asymptotic
minimum species richness estimators. Ecology 90:1125–1133.
Colwell, R.K. 2013. EstimateS: Statistical estimation of species richness and shared species
from samples, Version 9. Available online at http://purl.oclc.org/estimates. Accessed 29
December 2016.
Covell, C.V., Jr., and P.H. Freytag. 1979. The Malaise trap as a means of sampling butterfly
populations in Kentucky. Journal of the Lepidopterists’ Society 33:153–161.
Northeastern Naturalist Vol. 24, No. 1
K.W. McCravy
2017
23
Deans, A.M., J.R Malcolm, S.M. Smith, and M.I. Bellocq. 2005. Edge effects and the
responses of aerial insect assemblages to structural-retention harvesting in Canadian
boreal peatland forests. Forest Ecology and Management 204:249–266.
Di Giovanni, F., P. Cerretti, F. Mason, E. Minari, and L. Marini. 2015. Vertical stratification
of ichneumonid wasp communities: The effects of forest structure and life-history traits.
Insect Science 22:688–699.
Fraser, S.E.M., C. Dytham, and P.J. Mayhew. 2008. The effectiveness and optimal use
of Malaise traps for monitoring parasitoid wasps. Insect Conservation and Diversity
1:22–31.
Geroff, R.K., J. Gibbs, and K.W. McCravy. 2014. Assessing bee (Hymenoptera: Apoidea)
diversity of an Illinois restored tallgrass prairie: Methodology and conservation considerations.
Journal of Insect Conservation 18:951–964.
Ghahari, H., R.J. Lavigne, and F. Geller-Grimm. 2007. Bibliography of Asilidae (Insecta:
Diptera) 1996–2006. Far Eastern Entomologist 176:1–39.
Gleason, H.A., and A. Cronquist. 1991. Manual of Vascular Plants of Northeastern United
States and Adjacent Canada, 4th Edition. New York Botanical Garden, New York, NY.
910 pp.
Harris, R.J., and B.R. Burns. 2000. Beetle assemblages of Kahikatea forest fragments in a
pasture-dominated landscape. New Zealand Journal of Ecology 24:57–67.
Kartawich, L.M., and K.W. McCravy. 2010. A survey of the robber flies (Diptera: Asilidae)
of Ira and Reatha T. Post Wildlife Sanctuary, McDonough County, Illinois, USA. Transactions
of the Illinois State Academy of Science 103:31–38.
Kolbert, E. 2014. The Sixth Extinction: An Unnatural History. Henry Holt and Company,
New York, NY. 319 pp.
Larsen, M.N., and R. Meier. 2004. Species diversity, distribution, and conservation status
of the Asilidae (Insecta: Diptera) in Denmark. Steenstrupia 28:177–241.
Magurran, A.E. 2004. Measuring Biological Diversity. Blackwell Publishing, Malden, MA.
256 pp.
Malaise, R. 1937. A new insect-trap. Entomologisk Tidskrift 58:148–160.
McCravy, K.W., and K.A. Baxa. 2011. Diversity, seasonal activity and habitat associations
of robber flies (Diptera: Asilidae) in west-central Illinois. American Midland Naturalist
166:85–97.
McCravy, K.W., J.J. Bara, S.N. Hessler, L.K. Luxmore, K. S. Stinebaker, and S.E. Jenkins.
2009. Abundance and diversity of thread-waisted wasps (Hymenoptera: Sphecidae:
Sphecinae) at Alice L. Kibbe Life Science Station, Hancock County, Illinois, USA.
Transactions of the Illinois State Academy of Science 102:107–115.
McCravy, K.W., R.K. Geroff, and J. Gibbs. 2016. Malaise-trap sampling efficiency for
bees (Hymenoptera: Apoidea) in a restored tallgrass prairie. Florida Entomologist
99:321–323.
Ozanne, C.M.P. 2005. Sampling methods for forest understory vegetation. Pp. 58–76, In
S. Leather (Ed.). Insect Sampling in Forest Ecosystems. Blackwell Publishing, Malden,
MA. 303 pp.
Poole, R.W. 1996. Diptera. Pp. 15–604, In R.W. Poole and P. Gentili (Eds.). Nomina Insecta
Nearctica: A Check List of the Insects of North America, Vol. 3: Diptera, Lepidoptera,
Siphonaptera. Entomological Information Services, Rockville, MD. 1143 pp.
Raney, H. 2003. Robber flies. Available online at http://www.hr-rna.com/RNA/Rfly%20
pages/Robber%20list%20Illinois.htm. Accessed 29 December 2016.
Richards, L.A., and D.M. Windsor. 2007. Seasonal variation of arthropod abundance in gaps
and the understorey of a lowland moist forest in Panama. Journal of Tropical Ecology
23:169–176.
Northeastern Naturalist
24
K.W. McCravy
2017 Vol. 24, No. 1
Roberts, R.H. 1978. Effect of Malaise trap modifications on collections of Tabanidae. Mosquito
News 38:382–385.
Skillen, E.L., J. Pickering, and M.J. Sharkey. 2000. Species richness of the Campopleginae
and Ichneumoninae (Hymenoptera: Ichneumonidae) along a latitudinal gradient in eastern
North American old-growth forests. Environmental Entomology 29:460–466.
Topping, C.J. 1993. Behavioural responses of three linyphiid spiders to pitfall traps. Entomologica
Experimentalis et Applicata 68:287–293.
Townes, H. 1972. A light-weight Malaise trap. Entomological News 83:239–247.
Ulyshen, M.D., J.L. Hanula, and S. Horn. 2005. Using Malaise traps to sample ground
beetles (Coleoptera: Carabidae). The Canadian Entomologist 137:251–256.
Western Illinois University. 2016. Alice L. Kibbe Life Science Research Station. Available
online at http://www.wiu.edu/cas/biological_sciences/kibbe/index.php. Accessed 29
December 2016.
Wood, G.C. 1981. Asilidae. Pp. 549–573, In J.F. McAlpine, B.V. Peterson, G.E. Shewell,
H.J. Teskey, J.R. Vockeroth and D.M. Wood (Eds.). Manual of Nearctic Diptera. Vol. 1.
Research Branch, Agriculture Canada, Ottawa, ON, Canada. 674 pp.
Young, M. 2005. Insects in flight. Pp. 116–145, In S. Leather (Ed). Insect Sampling in Forest
Ecosystems. Blackwell Publishing, Malden, MA. 303 pp.