Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

Journal of Insect Conservation (2020) 24:765–777

https://doi.org/10.1007/s10841-020-00250-2

ORIGINAL PAPER

Insect diversity in vineyards, almond orchards, olive orchards, alfalfa


fields, and pastures in the San Joaquin Valley of California
Mark S. Sisterson1   · Donal P. Dwyer1 · Sean Y. Uchima1

Received: 29 April 2019 / Accepted: 10 June 2020 / Published online: 19 June 2020
© This is a U.S. government work and not under copyright protection in the U.S.; foreign copyright protection may apply 2020

Abstract
For many agricultural systems, limited data is available on abundance and diversity of insects that are not crop pests or their
natural enemies. As recent studies suggest that insect abundance and diversity is declining, there is a need to quantify insect
diversity within crop fields to determine what role crop fields play in maintaining diversity. In this study, alfalfa fields, almond
orchards, vineyards, olive orchards, and pastures located in the San Joaquin Valley of California were sampled. Ground sweep
samples were collected from all five habitats. In addition, foliar beat samples were collected from almond orchards, vine-
yards, and olive orchards. In total, ~ 240,000 arthropods were collected with the majority identified to family. Across crops
and sampling methods, 20 arthropod orders and 202 insect families were observed. Hemiptera was the most abundant order
of insect collected, representing an average of 61% of all arthropods collected. Diptera was the most diverse order of insect
collected, with a total of 59 dipteran families observed across crops. Of the 202 insect families observed, 85 were observed
in all 5 habitats (42%), whereas 48 families were observed in only one habitat (24%). Families observed in a single habitat
were often represented by only a few individuals. Principal component analysis indicated that the communities present in
the understory of vine and tree crops were more similar to each other than to the communities observed in pastures or alfalfa
fields. Much of the total insect biomass belonged to a few families that included known agricultural pests. In contrast, most
of the diversity was made up of families present in low to moderate abundance.

Keywords  Community ecology · Agroecosystem · Biological control · Pollination

Introduction diversity of insects that serve as natural enemies or pollina-


tors are common (e.g., Costello and Daane 1999; Benhadi-
Recent studies suggest that insect abundance and diversity Marin et al. 2011; Klein et al. 2012; Saunders et al. 2013;
has declined, with agricultural intensification hypothesized Shackelford et al. 2013). By comparison, fewer studies have
as one of many possible contributing factors (Hallmann et al. quantified diversity of non-pest herbivores or detritivores. As
2017; Leather 2018; Sánchez-Bayo and Wyckhuys 2019). a result, estimates of abundance and diversity of a consider-
While crop fields are routinely sampled to monitor abun- able portion of the insect community inhabiting crop fields
dance of economically damaging pests and beneficial insects is lacking for many agroecosystems.
(e.g., Landis et  al. 2000; Gurr et  al. 2017; Shackelford California has the largest agricultural economy in the
et al. 2013), insects with no clear economic value often go United States, with 6 of the top ten producing counties
uncounted. For example, studies quantifying abundance and located in the San Joaquin Valley (San Joaquin, Stanislaus,
Merced, Fresno, Tulare, and Kern Counties; CDFA 2017).
Electronic supplementary material  The online version of this Much of the agricultural land in the San Joaquin Valley is
article (https​://doi.org/10.1007/s1084​1-020-00250​-2) contains occupied by perennial crops that provide temporally stable
supplementary material, which is available to authorized users. habitats that are disrupted by pesticide applications. For
example, grapevines, almonds, and olives have stand life-
* Mark S. Sisterson
mark.sisterson@ars.usda.gov times that typically exceed 20 years (Klonsky et al. 1997;
Yaghmour et al. 2016; Fidelibus et al. 2018) and collec-
1
San Joaquin Valley Agricultural Sciences Center, USDA, tively cover more than 380,000 hectares in the San Joaquin
Agricultural Research Service, 9611 South Riverbend Valley (CDFA 2016, 2018a). Alfalfa fields and pastures are
Avenue, Parlier, CA 93648‑9757, USA

13
Vol.:(0123456789)

766 Journal of Insect Conservation (2020) 24:765–777

also common in the San Joaquin Valley, covering an area alternative site with the same crop was selected for sampling
exceeding 195,000 hectares (CDFA 2018b). Stand lifetimes in the second year. Over the course of the study, 5 vineyards,
of alfalfa fields are 3–5 years (Canevari and Putnam 2007) 4 almond orchards, 2 olive orchards, 4 alfalfa fields, and
and stand lifetimes for pastures are presumably measured 5 pastures were sampled, with sampling duration ranging
in years. Further, analysis of geographic information sys- between 9 and 24 months. Field sites were managed accord-
tems maps on crop distributions indicates that vineyards, ing to common grower practices. At the end of the study,
almonds, and alfalfa are frequently planted in proximity (i.e., growers were contacted and asked to provide information
within 1.5 km of each other) in the San Joaquin Valley of on application of insecticides, fungicides, and herbicides.
California (Sisterson et al. 2010). While pastures are spot-sprayed with herbicide (Forero et al.
In the case of tree and vine crops, the insect community 2015), insecticides are typically not applied. Accordingly,
consists of two portions: the portion inhabiting crop foliage pesticide records were not requested from pasture managers.
and the portion inhabiting plants growing in the understory For approximate locations of field sites and a description of
(e.g., Rieux et al. 1999; Simon et al. 2010). Ground cover neighboring crops see Supplementary Table S1. For pesti-
present on orchard or vineyard floors consists of weeds cide use history see Supplementary Table S2.
(Hanson et al. 2014) and in some cases cultivated ground
cover (e.g. Daane and Costello 1998). Presence of vege- Sampling
tation in the understory of tree and vine crops may affect
insect diversity in two ways. First, presence of ground cover In alfalfa fields and pastures, arthropods were collected by
increases plant diversity, which may affect insect diversity conducting 25 ground sweeps using a 40 cm diameter sweep
(e.g., Moreira et al. 2016). Second, as weed species found in net, with 2 to 3 sets of 25-sweeps collected per field on each
almond orchards, vineyards, and olive orchards are likely to sampling date. All collected arthropods and plant debris
be similar (Hanson et al. 2014), the communities present on were placed into a sealable plastic bag and returned to the
ground cover in tree and vine crops may be similar. laboratory for processing. In vineyards, almond orchards,
To describe the insect community found in the San and olive orchards samples were collected using two meth-
Joaquin Valley agroecosystem, alfalfa fields, almond ods. First, 25 sweeps of ground cover present in almond
orchards, vineyards, olive orchards, and pastures were sam- orchards, vineyards, and olive orchards were collected as
pled. The approach used here was to intensively sample a was done in alfalfa fields and pastures. At each sampling
limited number of sites of each crop and identify all col- location, the sampler began at the edge of the vineyard or
lected insects to the family level. By intensively sampling orchard and moved inward ~ 1 m with each pendulum swing
sites, the goal was to account for all families commonly of the net. Specifically, 6–8 sets of 25-sweeps were collected
observed in these five crops based on the sampling meth- per orchard or vineyard on each sampling date, with sweep
ods used. Emphasis was placed on documenting presence locations uniformly distributed across each side of the field.
and abundance of families that are not considered pests Ground sweeps were not collected if ground cover was
and whose contribution to ecosystem services are poorly absent due to tillage or herbicide treatment or too short to
described. Describing the insect community present in the sweep. For statistical purposes, ground sweeps not collected
San Joaquin Valley agroecosystem is important for two rea- due to absence of ground cover were treated as zeros as the
sons. First, surveys provide baseline data for comparison insects being sampled are unlikely to inhabit bare ground.
to future studies. Second, surveys may benefit agricultural In contrast, ground sweeps that were not collected due to
production by documenting presence of insects that provide ground cover being too short to sweep were treated as miss-
ecosystems services affecting crop yield, but whose contri- ing values as no assessment could be made.
butions are overlooked. To directly sample almond, grape, and olive foliage, beat
samples were collected. Foliage beats were collected by
beating the foliage with a stick while holding the sweep net
Materials and methods beneath the crop canopy. The sampler started at the edge of
the orchard or vineyard and beat the foliage ten times while
Field sites walking into the vineyard or orchard. Beats of foliage were
taken on each visit to a vineyard, almond orchard, and olive
Field sites were located in Fresno and Tulare counties, orchard, with 6–8 sets of beat samples collected per orchard
California, USA (Table 1). In 2012, 4 vineyards, 3 almond or vineyard on each sampling date. Beat sample locations
orchards, 2 olive orchards, 3 alfalfa fields, and 4 pastures were uniformly distributed across each side of the field. In
were selected for sampling. Sites were sampled for 2 years winter when grapevines and almond trees are dormant, the
or until a site was removed from production. For sites that content of the sweep net was bagged only if arthropods were
were removed from production at the end of the first year, an present. Sampling began in January of 2012 and sweep and

13
Journal of Insect Conservation (2020) 24:765–777 767

Table 1  Description of field Cropa Cultivar(s) Date planted Period sampled


sites and sampling duration
Grape
­A1 French colombard and Zinfandel 1985 2012-Jan to 2013-Mar
­A2 Cabernet Sauvignon and Rhone types 1993/1998 2013-Apr to 2013 Dec
C Thompson seedless 1979/2011 2012-Jan to 2013-Dec
D Red Globe 1996 2012-Jan to 2013-Dec
E Thompson seedless 1962 2012-Jan to 2013-Sep
Almond
B Nonpareil, Aldridge 2010/2011 2012-Jan to 2013-Dec
C Nonpareil, Monterey 2011/2012 2012-Jan to 2013-Dec
D Nonpareil, Aldridge, Monterey 2007 2012-Jan to 2013-Dec
F Nonpareil, Sonora, Carmel 1987 2013-Feb to 2013-Dec
Olive
A Koroneiki, Arbequina, Arbosona 2003 2012-Jan to 2013-Nov
B Arbequina, Arbosona 2009 2012-Jan to 2013-Dec
Alfalfa
A Roundup ready NK 2011 2012-Jan to 2013-Jun
B 57N58 2010 2012-Feb to 2012-Dec
E Pioneer 5728 2009 2012-Jan to 2013-Dec
F Dairyland Magnum 80 2012 2013-Jan to 2013-Dec
Pasture
A Grasses NA 2012-Jan to 2013-Dec
B Grasses NA 2012-Jan to 2013-Dec
­C1 Grasses NA 2012-Jan to 2013-Mar
­C2 Grasses NA 2013-Apr to 2013-Dec
D Grasses NA 2012-Jan to 2012-Dec
a
 Crops with the same letter were < 2 km apart. Distances between sites with different letters were substan-
tially > 2 km

beat samples were collected monthly until the end of the Samples are stored at the San Joaquin Valley Agricultural
study in December of 2013. In total ~ 2100 sweep/beat sam- Sciences Center in Parlier, CA.
ples were collected, resulting in the collection of ~ 240,000
arthropods. Analysis

Sample processing Average abundance per sweep or beat sample was deter-
mined for each insect family. As our goal was to document
On return to the laboratory all samples were placed into presence and distribution across crops, principal component
a −20 °C freezer. To process each sample, the contents analyses were used to identify latent variables that distin-
were emptied into a 15 cm diameter petri dish and exam- guished between the insect communities observed in each
ined under a dissecting microscope. Non-insect arthropods crop at the order and family level. JMP stats (SAS, Cary,
were identified to order (Acari and Araneae). Adult and NC) was used to complete principal component analyses.
juvenile insects were identified to family using Triplehorn The intent was to intensively sample a limited num-
and Johnson (2005). Some juvenile insects can be difficult ber of sites to determine total families present based on
to identify to family (e.g., some Diptera, Lepidoptera, and the sampling methods used. To determine if sampling
Coleoptera larvae). For such specimens, identification was achieved this goal, rarefaction was conducted. If rarefac-
limited to order. In total, 98% of insects were identified to tion curves approach a horizontal asymptote, there is con-
family. Ideally, Thysanoptera, Collembola, and Pscoptera fidence that most insect families that may be captured
should be slide mounted for identification which was not via ground sweeps or foliage beats were observed. Con-
practical given the quantity collected. Accordingly, family ducting rarefaction also provides the opportunity to com-
identification of these three orders is tentative. After iden- pare diversity among crops while controlling for num-
tification all arthropods were preserved in 70% ethanol. ber of insects observed (Gotelli and Colwell 2001). The

13

768 Journal of Insect Conservation (2020) 24:765–777

rarefaction program was coded in C++ (Microsoft Visual


C++; Redmond, WA). For each run of the rarefaction
program, a specified number of arthropods was randomly
selected from the pool of all arthropods observed in each
habitat without replacement and number of families
observed recorded. For each specified number of arthro-
pods, 1000 repetitions of the rarefaction program were
completed and 95% confidence intervals determined.
To compare number of families observed in each order
among crops, an approach based on rarefaction was used to
control for differences in number of insects collected in each
crop. To accomplish this, number of families in each order
collected via ground sweeps or foliage beats was estimated
for a standardized number of arthropods. For ground sweep
samples, number of families in each order was estimated
by randomly selecting 5000 arthropods from the pool of
all arthropods collected via grounds sweeps in each crop
and determining number of families observed in each order.
The process was repeated 1000 times and mean number of
families observed per order was determined and 95% con-
fidence intervals constructed. For beat samples, number of
families per order was determined by randomly selecting
1700 arthropods.

Results

Description of the arthropod community in ground


sweep samples at the level of order

A principal component analysis of ground sweep samples


conducted at the order level resulted in 6 components with
eigenvalues > 1 (Fig. 1a). Across crops, 20 arthropod orders
were observed in ground sweep samples, with 12 orders
observed in all habitats (Fig. 2a; Hemiptera, Diptera, Thy-
sanoptera, Hymenoptera, Coleoptera, Aranea, Lepidoptera,
Collembola, Pscoptera, Orthroptera, Neuroptera, and Acari).
Eight orders were not observed in all habitats (Fig. 2a; Man-
todea, Odonata, Ephemeroptera, Dermaptera, Blatodea, Fig. 1  Principal component score plots for ground sweeps (a) and
Isopoda, Embidiina, and Strepsiptera). foliage beats (b) conducted at the order level. For ground sweeps,
Score plots of components 1 and 2 provided the most component 1 had an eigenvalue of 7.3 and accounted for 36.6% of the
informative separation of ground foliar communities at variation among sampling locations and component 2 had an eigen-
value of 3.3 and accounted for 16.7% of the variation among sam-
the level of order (Fig.  1a, Supplementary Table  S3). pling locations. For foliage beats, component 1 had an eigenvalue of
Ten orders had loadings on component 1 that were > 0.5 4.1 and accounted for 25.4% of the variation among sampling loca-
(Fig. 2a; Hemiptera, Diptera, Thysanoptera, Hymenoptera, tions and component 2 had an eigenvalue of 3.3 and accounted for
Coleoptera, Araneae, Lepidoptera, Psocoptera, Neuroptera, 20.6% of the variation among sampling locations. Orders with load-
ings > 0.50 on component 1 or 2 are shown in Fig.  2. Sampling of
and Odonata). As observations from alfalfa fields had the almond orchards, olive orchards, vineyards, pastures, and alfalfa
greatest component 1 scores (Fig. 1a), the ten orders with fields during 2012–2013
high loadings on component 1 were observed in greater
abundance in alfalfa fields compared to the four other sam-
pled habitats (Fig. 2a). In fact, component 1 was correlated in the community at the order level between alfalfa fields
with total arthropod abundance (F = 85.3, df = 1, 18, P < and the other sampled crops was abundance. Component 1
0.0001, r2 = 0.85), indicating that the primary difference and 2 scores were similar for vineyards, almond orchards,

13
Journal of Insect Conservation (2020) 24:765–777 769

Dermaptera, and Odonata) were observed in one or two


of the three habitats. Orders that were not observed across
all three habitats were generally found in low abundance
(Fig. 2b).
Score plots of components 1 and 2 indicated minor differ-
ences in community composition at the order level (Fig. 1b,
Supplementary Table S4). Observations from olive orchards
had the greatest component 1 scores. Four orders had posi-
tive loadings on component 1 that were > 0.5 (Fig. 2b;
Acari, Collembola, Mantodea, and Odonata), suggesting
greater abundance in olive orchards. Three orders had nega-
tive loadings < − 0.5 on component 1 (Fig. 2b; Coleop-
tera, Dermaptera, and Ephemeroptera), suggesting lower
abundance in olive orchards. Observations from vineyards
had greater component 2 scores than almond orchards. Five
orders had positive loadings on component 2 that were >
0.50 (Fig. 2b; Aranea, Diptera, Hemiptera, Lepidoptera, and
Orthoptera) suggesting greater abundance in vineyards than
in almond orchards. Thysanoptera had a negative loading on
component 2 that was < -0.50 suggesting greater abundance
in almond orchards than in vineyards (Fig. 2b).

Insect diversity at the family level in ground sweeps


and foliage beat samples

Across beat and sweep samples, 202 insect families were


observed, with 85 families (42%) observed in all five sam-
pled habitats. In ground sweeps, 18% of families observed in
each crop were represented by a single individual (Fig. 3a).
In foliage beats, 32% of families observed in each crop were
Fig. 2  Mean (± SE) abundance of each arthropod order per sample represented by a single individual (Fig. 3b). Averaged across
for ground sweeps (a) and foliage beats (b). For ground sweeps, a crops, 5% of families were represented by > 750 individu-
sample consisted of 25 pendulum swings of the net. For foliage beats, als in ground sweeps and 1% of families were represented
a sample consisted of ten beats of the foliage. Numbers above bars by > 750 individuals in beats of foliage (Fig. 3). Families
indicates that an order had a loading of > 0.50 on principal compo-
nent 1 or 2 (see Fig. 1). Sampling of almond orchards, olive orchards, observed in high abundance included insects that are eco-
vineyards, pastures, and alfalfa fields during 2012–2013 nomically important pests (e.g., Aphididae, Cicadellidae,
Membracidae, Miridae, Thysanoptera, Curculionidae) and
natural enemies (e.g., Anthocoridae and Braconidae). Most
olive orchards, and pastures, suggesting little difference in of the insect biomass observed belonged to a few families,
abundance of each order (Figs. 1a, 2a). whereas most of the diversity was comprised of families pre-
sent in low to moderate abundance (Fig. 3). The aforemen-
Description of the arthropod community in foliage tioned pattern is typical of a wide range of systems (McGill
beat samples at the level of order et al. 2007).
Rarefaction curves for ground sweep and foliage beats
A principal component analysis on foliage beat samples con- were decelerating (Fig. 4), suggesting that common families
ducted at the order level had 6 components with eigenvalues were observed, and a few rare families were not observed.
> 1 (Fig. 1b). Across vineyards, almond orchards, and olive On a per individual basis, family diversity in ground sweep
orchards, 16 arthropod orders were observed in foliage beat samples was greatest in pastures, intermediate in vineyards,
samples. Ten orders were observed in beats samples col- almond orchards, and olive orchards, and lowest in alfalfa
lected from all three habitats (Fig. 2b; Hemiptera, Diptera, fields (Fig. 4a). As rarefaction measured diversity on a per
Thysanoptera, Hymenoptera, Araneae, Coleoptera, Neurop- individual basis, family diversity in alfalfa fields appeared
tera, Psocoptera, Collembola, and Lepidoptera). Six orders low as Aphididae represented 50% of all arthropods col-
(Fig.  2b; Acari, Mantodea, Orthoptera, Ephemeroptera, lected in alfalfa. Comparison of rarefaction curves for beat

13

770 Journal of Insect Conservation (2020) 24:765–777

Fig. 4  Rarefactions curves indicating the number of insect families


observed based on the number of arthropods collected for ground
sweeps (a) and foliage beats (b). Error bars represent 95% confidence
intervals. Sampling of almond orchards, olive orchards, vineyards,
Fig. 3  Histogram of the number of individuals observed per insect pastures, and alfalfa fields during 2012–2013
family for ground sweeps (a) and foliage beats (b). Sampling of
almond orchards, olive orchards, vineyards, pastures, and alfalfa
fields during 2012–2013. Bin ranges were as follows: 1 (1), 5 (2 to for Hymenoptera, Hemiptera, and Coleoptera. For ground
27), 50 (28 to 275), 500 (276 to 750), > 750 (> 750) sweep samples, an average of 17 hymenopteran, 15 hemip-
teran, and 10 coleopteran families were observed in a ran-
dom sample of 5000 arthropods. For foliage beat samples, an
samples suggested that family diversity was greatest on olive average of 14 hymenopteran, 15 hemipteran, and 8 coleop-
foliage, intermediate on almond foliage, and least on grape teran families were observed in a random sample of 1700
foliage. One a per individual basis, family diversity on grape arthropods. The remaining orders were represented by an
foliage appeared low as 58% of all arthropods collected in average of 3 or fewer families in random samples of 5000 or
beats of grape foliage belonged to the Cicadellidae. 1700 arthropods for sweep and beat samples, respectively.
Patterns of family diversity within orders was simi-
lar for ground sweep and foliar beat samples and similar Description of the insect community in ground
across crops (Fig. 5). Across crops, the greatest diversity sweep samples at the level of family
was observed in the Diptera with an average of 32 dipteran
families in a random sample of 5000 arthropods collected A principal component analysis conducted at the fam-
via ground sweeps and an average of 26 dipteran families ily level for ground sweep samples had 18 components
observed in a random sample of 1700 arthropods collected with eigenvalues > 1. Score plots of component 1 and
via foliage beats. Diversity at the family level was moderate component 2 were most informative (Fig. 6a). In total, 65

13
Journal of Insect Conservation (2020) 24:765–777 771

Fig. 6  Principal component score plots for ground sweeps (a) and


foliage beats (b) conducted at the Family level. For ground sweeps,
Fig. 5  Number of insect families observed in each order for ground component 1 had an eigenvalue of 39.9 and accounted for 20.7% of
sweeps (a) and foliage beats (b). To eliminate effects of insect abun- the variation among sampling locations and component 2 had an
dance on family diversity, number of families observed in each order eigenvalue of 22.3 and accounted for 11.5% of the variation among
was estimated by randomly selecting 5000 arthropods from sweep sampling locations. For foliage beats, component 1 had an eigenvalue
sample data sets for each crop or 1700 arthropods from beat sample of 27.5 and accounted for 20.9% of the variation among sampling
data sets for each crop and determining number of families observed locations and component 2 had an eigenvalue of 22.6 and accounted
in each order. This process was repeated 1000 times and mean num- for 17.1% of the variation among sampling locations. Families in
ber of families observed in each order was determined. Error bars the Diptera, Hymenoptera, Hemiptera, and Coleoptera with loadings
represent 95% confidence intervals. Orders represented by a single having absolute values > 0.50 on components 1 or 2 are shown in
family were omitted from the figure (e.g., Dermaptera: Forficulidae, Figs.  7, 8, 9, and 10. Sampling of almond orchards, olive orchards,
Mantodea: Mantidae etc.) as were orders that were sufficiently rare vineyards, pastures, and alfalfa fields during 2012–2013
as to not typically be observed in the 5000 or 1700 arthropod sam-
ple (e.g., Ephemeroptera, Odonata, Strepsiptera). Sampling of almond
orchards, olive orchards, vineyards, pastures, and alfalfa fields during
2012–2013

13

772 Journal of Insect Conservation (2020) 24:765–777

families had loadings on component 1 with an absolute positive loadings on component 1 were generally in greater
value > 0.50 and 27 families had loadings on component abundance in alfalfa fields than in the other sampled crops
2 with an absolute value > 0.50 (for loadings see Sup- (Figs. 7, 8, 9, 10). Likewise, as pastures had the largest
plementary Table S5; for raw counts see Supplementary component 2 scores (Fig. 6a), families with high positive
Table  S6). As observations from alfalfa fields had the loadings on component 2 were generally found in greater
largest component 1 scores (Fig. 6a), families with high abundance in pastures than in the other sampled habitats

Fig. 7  Mean (± SE) abundance of families belonging to Diptera in that a family had a loading on component 1 or 2 that had an abso-
beat or sweep samples. For sweep samples (a), families with an aver- lute value > 0.50 (see Fig. 6). Positive values indicate positive load-
age abundance across crops of < 0.05 per sample were omitted. For ings, negative values indicate negative loadings. Sampling of almond
beat samples (b), families with an average abundance across crops orchards, olive orchards, vineyards, pastures, and alfalfa fields during
of < 0.002 per sample were omitted. Numbers above bars indicates 2012–2013

13
Journal of Insect Conservation (2020) 24:765–777 773

Fig. 9  Mean (± SE) abundance of families belonging to Hemiptera in


Fig. 8  Mean (± SE) abundance of families belonging to Hymenop- beat or sweep samples. For sweep samples (a), families with an aver-
tera in beat or sweep samples. For sweep samples (a), families with age abundance across crops of < 0.05 per sample were omitted. For
an average abundance across crops of < 0.05 per sample were omit- beat samples (b), families with an average abundance across crops
ted. For beat samples (b), families with an average abundance across of < 0.002 per sample were omitted. Numbers above bars indicates
crops of < 0.002 per sample were omitted. Numbers above bars that a family had a loading on component 1 or 2 that had an abso-
indicates that a family had a loading on component 1 or 2 that had lute value > 0.50 (see Fig. 6). Positive values indicate positive load-
an absolute value > 0.50 (see Fig.  6). Positive values indicate posi- ings, negative values indicate negative loadings. Sampling of almond
tive loadings, negative values indicate negative loadings. Sampling orchards, olive orchards, vineyards, pastures, and alfalfa fields during
of almond orchards, olive orchards, vineyards, pastures, and alfalfa 2012–2013
fields during 2012–2013

habitats (Fig. 8a). Fifteen hymenopteran families had load-


(Figs. 7, 8, 9, 10). Component 1 and 2 scores were simi- ings with an absolute value > 0.50 on component 1 and 6
lar for observations from vineyards, almond orchards, and families had loadings with absolute values > 0.50 on com-
olive orchards, indicating that the communities present in ponent 2. A total of 30 hemipteran families were observed
the understory of vine and tree crops were more similar to across crops, with 13 families (43%) observed in all five
each other than to the communities observed in pastures sampled habitats (Fig. 9a). Twelve hemipteran families had
or alfalfa fields (Fig. 6a). loadings with an absolute value > 0.50 on component 1
A total of 55 dipteran families were observed across and 3 families had loadings with an absolute value > 0.50
crops, with 29 families (53%) observed in all five sampled on component 2. A total of 30 coleopteran families were
crops (Fig. 7a). A total of 23 dipteran families had loadings observed, with 8 families (27%) observed in all five sampled
with an absolute value > 0.50 on component 1 and 9 families habitats (Fig. 10a). In total, 7 families had loadings with
had loadings with absolute values > 0.50 on component 2. absolute values > 0.50 on component 1 and 4 families had
A total of 41 hymenopteran families were observed across loadings with absolute values > 0.50 on component 2.
crops, with 15 families (37%) observed in all five sampled

13

774 Journal of Insect Conservation (2020) 24:765–777

on component 2 (for loadings see Supplementary Table S7;


for raw counts see Supplementary Table S8).
Across crops, 40 families of Diptera were observed in
beat samples, with 19 dipteran families (46%) observed in
beat samples from all three crops (Fig. 7b). Sixteen dipteran
families had loadings with absolute values > 0.5 on com-
ponent 1 and 8 dipteran families had loadings with absolute
values > 0.50 on component 2. A total of 26 hymenopteran
families were observed in beats samples across crops, with
8 families (31%) observed in samples from all three crops
(Fig. 8b). In total, 3 families of Hymenoptera had loadings
with absolute values > 0.5 on component 1 and 3 families
had loadings with absolute values > 0.5 on component 2.
Across crops, 21 hemipteran families were observed, with
12 families (55%) observed in all three crops (Fig. 9b). Five
hemipteran families had loadings with absolute values >
0.50 on component 1 and 4 hemipteran families had loadings
with absolute values > 0.50 on component 2. Across crops,
16 families of Coleoptera were observed in beat samples,
with 5 families (31%) observed in all three crops (Fig. 10b).
Four coleopteran families had loadings with absolute values
> 0.50 on component 1 and one coleopteran family had a
loading with an absolute value > 0.50 on component 2.

Discussion

There is increasing recognition that adoption of cropping


practices that conserve insect diversity may benefit crop
production through improved biological control, pollination
Fig. 10  Mean (± SE) abundance of families belonging to Coleop- services, and decomposition/nutrient cycling (e.g., Simon
tera in beat or sweep samples. For sweep samples (a), families with et al. 2010; Kleijn et al. 2011; Lichtenberg et al. 2017; Win-
an average abundance across crops of < 0.05 per sample were omit-
ted. For beat samples (b), families with an average abundance across
ter et al. 2018). Designing strategies that conserve insect
crops of < 0.002 per sample were omitted. Numbers above bars indi- diversity within agricultural systems requires an under-
cates that a family had a loading on component 1 or 2 with an abso- standing of community structure at the agroecosystem level.
lute value > 0.50 (see Fig. 6). Positive values indicate positive load- Here, the insect communities found in alfalfa fields, almond
ings, negative values indicate negative loadings. Sampling of almond
orchards, olive orchards, vineyards, pastures, and alfalfa fields during
orchards, vineyards, olive orchards, and pastures located
2012–2013 in the San Joaquin Valley of California were described. A
total of 205 insect families were observed, with 85 families
observed in all 5 habitats. The insect communities present in
Description of the insect community in foliage beat the five sampled habitats had several similarities and some
samples at the level of family differences.
Members of the Hemiptera were the most abundant
A principal component analysis conducted at the family across crops and sampling methods. High abundance of
level for samples collected via foliage beats from vineyards, Hemiptera was primarily due to presence of herbivores,
almond orchards, and olive orchards required 10 compo- particularly from the families Aphididae and Cicadelli-
nents to explain 100% of the variance among observations. dae. Diptera was the second most abundant insect order
Component 1 scores tended to be greater in vineyards than representing 22% of all arthropods observed and harbored
in almond orchards and olive orchards, whereas component the greatest diversity at the family level. High relative
2 scores tended to be greater in olive orchards than in vine- abundance and/or diversity of the Diptera at the family
yards and almond orchards (Fig. 6b). In total, 33 families level has been observed in numerous cropping systems
had loadings with an absolute value > 0.50 on component 1 (Sáenz-Romo et al. 2019; Ruano et al. 2004; Prischmann
and 24 families had loadings with an absolute value > 0.50 et al. 2005; Ponce et al. 2011; Poveda et al. 2012; Pocock

13
Journal of Insect Conservation (2020) 24:765–777 775

and Jennings 2008). Despite high relative contribution of of Diptera visited almond flowers, to our knowledge, the
Diptera to total insect biomass and diversity in agricul- only dipteran family that has been shown to successfully
tural systems, the role of Diptera in agricultural systems pollinate almonds are members of the Syrphidae (Henselek
is understudied. For example, a metanalysis examin- et al. 2018).
ing effects of diversified farming on insect diversity by One approach to conserving diversity in orchard and vine
Lichtenberg et al. (2017) reviewed 62 studies, with only crops is maintenance of plants in the understory (e.g., Daane
15 studies quantifying abundance of at least a single dip- and Costello 1998; Daane et al. 2018; Winter et al. 2018).
teran family. Further, a recent review on the decline of Here, a diverse group of insects was observed on weeds nat-
the entomofauna by Sánches-Bayo and Wyckhuys (2019) urally occurring in the understory of vine and orchard crops.
implies that few studies have been conducted on members Principal component analyses suggested that the insect com-
of the Diptera with studies on Diptera primarily focusing munities present on weeds in the understory of vineyards,
on a single family (Syrphidae). almond orchards, and olive orchards were similar. As weed
Natural enemies including spiders (Araneae), lacewings species present in vineyards, almond orchards, and olive
(Neuroptera: Chysopidae), damsel bugs (Hemiptera: Nabi- orchards are likely to be similar (Hanson et al. 2014), it is
dae), minute pirate bugs (Hemiptera: Anthocoridae), big reasonable to expect the community inhabiting those weeds
eyed bugs (Hemiptera: Geocoridae), ladybeetles (Coleop- to be similar. Rarefaction indicated that on a per individual
tera: Coccinellidae), and parasitoid wasps (e.g., Hyme- basis family richness in sweeps of ground cover was greater
noptera: Braconidae, Ichneumonidae, Eulophidae, and in pastures, vineyards, almond orchards, and olive orchards
Pteromalidae) were common and observed in all habitats. compared to alfalfa fields. As an increase in plant diver-
Identification of spiders was limited to order and there is sity is anticipated to increase insect diversity (e.g., Moreira
likely considerable diversity within the Araneae in the San et al. 2016), higher family richness on a per individual basis
Joaquin Valley agroecosytem. Costello and Daane (1999) may be explained by greater plant diversity in pastures and
observed 6 families of spiders and 5 families of predaceous in the understory of vineyards, almond orchards, and olive
insects in Californian vineyards. Benhadi-Marin et al. (2011) orchards compared to alfalfa fields. Nonetheless, arthro-
observed 14 families of spiders in almond orchards in Por- pod abundance was greatest in alfalfa fields and abundance
tugal. Many of the natural enemy families observed in high of families common to all crops was generally greatest in
abundance are recognized as important natural enemies in alfalfa fields. Accordingly, it is possible that alfalfa fields
agricultural systems. By comparison, the role of predatory play an important role in the persistence of much of the
Diptera has received less attention (Pfister et al. 2017). Here, diversity present among crops. In fact, alfalfa has previously
relatively high abundance of five families of predacious/par- been referred to as a ‘field insectary’ which produces large
astic Diptera was observed (Dolichopodidae, Empididae, numbers of generalist predators and parasitoids (Summers
Tachinidae, Syrphidae, and Pipunculidae). Soares et  al. 1998).
(2016) also observed relatively high abundance of Dolicho- In this study, ~ 240,000 arthropods were collected, with
podidae in an experimental vineyard in Brazil and Rieux the majority identified to family. As sampling methodology
et al. (1999) reported that Empididae were common in sam- affects the types of insects captured, alternative sampling
ples collected from ground cover in pear orchards in France. methods may yield insect families that were not observed.
In this study, almond was the only crop that required an For example, pitfall trapping is likely to yield families
insect pollinator (Yaghmour et al. 2016). The only fam- that live in close association with the soil surface that are
ily of native bees observed in almond orchards during our unlikely to be captured by sweep net. Similarly, insects that
study belonged to the Halictidae and observations of Hal- infest fruit, bore into plants, or make galls may be under-
ictidae occurred outside of bloom. Many species of Dip- represented in ground sweeps or foliar beats. Thus, our
tera are associated with flower and may serve as pollinators assessment quantifies insects found on the foliage of ground
(Ssymank et al. 2008; Rader et al. 2016). Several of the cover and crop vegetation. In addition, the field sites used in
dipteran families observed in almond orchards during this this study were in areas with at least two other crops in the
study are associated with flowers (e.g. Syrphidae, Heleo- vicinity. Analysis of GIS maps indicates that crop diversity
myzidae, Empididae, Drosophilidae, Anthomyiidae, Chloro- varies throughout the San Joaquin Valley (e.g., Sisterson
pidae, Ceratopogonidae, Cecidomyiidae, Muscidae; Larson et al. 2008, 2010). As landscape effects are documented to
et al. 2001; Wardhaugh 2015; Hahn and Brühl 2016). Two influence insect abundance (e.g., Shackelford et al. 2013;
observational studies on almond pollination determined Pfister et al. 2017), sampling in areas with higher or lower
that 17.4% of visits to almond flowers in California (Klein crop diversity may be associated with changes in the insect
et al. 2012) and 4.1% of visit to almond flowers in Mallorca communities in each crop.
(Spain; Alomar et al. 2018) were by members of the Diptera. For identification, the keys provided by Triplehorn and
While Alomar et al. (2018) documented that 12 families Johnson (2005) were used. Identification of many insect

13

776 Journal of Insect Conservation (2020) 24:765–777

families is straightforward. However, in some cases it can canopy in northeaster Portugal: a methodological approach. Ento-
be difficult to distinguish between closely related families, mol Sci 14:347–358
Canevari M, Putnam DH (2007) Managing depleted alfalfa stands:
particularly if antenna, tarsi, or wings are damaged. Given overseeding and other options. In: Summers CG, Putnam DH (eds)
the quantity of insects identified (~ 240,000 arthropods were Irrigated alfalfa management for Mediterranean and Desert Zones.
processed), it is unrealistic to assume that no misidentifica- University of California Agriculture and Natural Resources, Davis
tions were made. If insects were misidentified, they were [CDFA] California Department of Food and Agriculture (2016) Cali-
fornia almond acreage report. California Department of Food and
likely placed in a closely related family (i.e., within the same Agriculture, Sacramento
superfamily). In addition, Triplehorn and Johnson (2005) [CDFA] California Department of Food and Agriculture (2017) Cali-
may not include recent taxonomic revisions. For example, fornia agricultural statistics review 2016–2017. California Depart-
Triplehorn and Johnson (2005) recognize only two families ment of Food and Agriculture, Sacramento
[CDFA] California Department of Food and Agriculture (2018a) Grape
of Diptera within the superfamily Empidoidea: Empididae acreage report 2017 crop. California Department of Food and
and Dolichopodidae, whereas Moulton and Wiegmann Agriculture, Sacramento
(2007) provides support for dividing the superfamily into [CDFA] California Department of Food and Agriculture (2018b)
five families (Moulton and Wiegmann 2007). California county agricultural commissioners’ reports crop year
2015–2016. California Department of Food and Agriculture,
Agricultural fields are some of the most extensively Sacramento
sampled habitats for insects, although sampling typically Costello MJ, Daane KM (1999) Abundance of spiders and insect preda-
targets a limited number of economically important taxa. tors on grapes in Central California. J Arachnol 27:531–538
Thus, life histories and ecosystem services provided by a Daane KM, Costello MJ (1998) Can cover crop reduce leafhopper
abundance in vineyards? Calif Agricult 52:27–33
few taxa found in agroecosystems are often well described Daane KM, Hogg BN, Wilson H, Yokota GY (2018) Native grass
with much less information available for the remainder of ground covers provide multiple ecosystems services in Californian
the insect community. Given concerns about the decline of vineyards. J Appl Ecol 55:2473–2483
insects worldwide, documenting insect diversity in agricul- Fidelibus M, El-kereamy A, Haviland D, Hembree K, Zhuang G, Stew-
art D, Sumner DA (2018) Sample costs to establish and produce
tural areas is important in determining what role agricul- table grapes: San Joaquin Valley—South. University of California
tural fields may play in conserving diversity. As the human Cooperative Extension, Oakland
population continues to increase, there is a need for research Forero LC, Ingram R, Davy J, Nader G, Klonsky K, Stewart D (2015)
that focuses on improving crop yields. The challenge for Sample costs to produce pasture: Sacramento Valley. University
of California Cooperative Extension, Oakland
agricultural entomologists is to identify insect conservation Gotelli NJ, Colwell RK (2001) Quantifying biodiversity: procedures
strategies that improve crop yields by leveraging the ecosys- and pitfalls in the measurement and comparison of species rich-
tem services provided by insects. ness. Ecol Lett 4:379–391
Gurr GM, Wrateen SD, Landis DA, You M (2017) Habitat management
Acknowledgements  Funding for this work was from United States to suppress pest populations: progress and prospects. Annu Rev
Department of Agriculture (USDA) Agricultural Research Service Entomol 62:91–109
appropriated Project 2034-22000-012-00D. Mention of trade names Hahn M, Brühl CA (2016) The secret pollinators: an overview of moth
or commercial products in this publication is solely for the purpose of pollination with a focus on Europe and North America. Arthropod
providing specific information and does not imply recommendation or Plant Interact 10:21–28
endorsement by the USDA. USDA is an equal opportunity employer. Hallmann CA, Sorg M, Jongejans E, Siepel H, Hofland N, Schwan
H, Stenmans W, Müller A, Sumser H, Hörren T, Goulson D,
de Kroon H (2017) More than 75 percent decline over 27 years
Compliance with ethical standards  in total flying insect biomass in protected areas. PLoS ONE
12:e01858109
Conflict of interest  The authors declare that the they have no conflict Hanson BD, Rocoroni J, Hembree KJ, Molinar R, Elmore CL (2014)
of interest. Tree, vine, and soft-fruit crops. In: Fennimore SA, Bell CE (eds)
Principals of weed control. California Weed Science Society, Sali-
Ethical approval  This article does not contain any studies with human nas, pp 334–389
participants or vertebrate animals. All authors have approved the manu- Henselek Y, Eilers EJ, Kremen C, Hendrix SD, Klein A-M (2018)
script and agree to submission to Journal of Insect Conservation. Pollination requirements of almond (Prunus dulcis): combing
laboratory and field experiments. J Econ Entomol 111:1006–1013
Kleijn D, Rundlöf M, Scheper J, Smith HG, Tscharntke T (2011) Does
conservation on farmland contribute to halting the biodiversity
decline? Trends Ecol Evol 26:474–481
References Klein A-M, Brittain C, Hendrix SD, Thorp R, Williams N, Kremen
C (2012) Wild pollination services to California almond rely on
Alomar D, González-Estévez MA, Traveset A, Lázaro A (2018) The semi-natural habitat. J Appl Ecol 49:723–732
intertwined effects of natural vegetation, local flower community, Klonsky K, Sibbett GS, Freeman M, Livingston P (1997) Sample costs
and pollinator diversity on the production of almond trees. Agric to establish a manzanillo olive orchard and produce olives in the
Ecosyst Environ 264:34–43 southern San Joaquin Valley. University of California Cooperative
Benhadi-Marin J, Pereira JA, Barrientos JA, Bento A, Santos SAP Extension, Oakland
(2011) Diversity of predaceous arthropods in the almond tree

13
Journal of Insect Conservation (2020) 24:765–777 777

Landis DA, Wratten SD, Gurr GM (2000) Habitat management to con- Sáenz-Romo MG, Veas-Bernal A, Martínez-García H, Campos-Herrera
serve natural enemies or arthropod pests in agriculture. Annu Rev RC, Ibáñez-Pascual S, Martínez-Villar E, Pérez-Moreno I, Marco-
Entomol 45:175–201 Mancebón VS (2019) Ground cover management in a Mediter-
Larson BMH, Kevan PG, Inouye DW (2001) Flies and flowers: taxo- ranean vineyard: impact on insect abundance and diversity. Agric
nomic diversity of anthophiles and pollinators. Can Entomol Ecosyst Environ 283:106571
133:439–465 Sánchez-Bayo F, Wyckhuys KAG (2019) Worldwide decline of the
Leather SR (2018) “Ecological armageddon"—more evidence for the entomofauna: a review of its drivers. Biol Conserv 232:8–27
drastic decline in insect numbers. Ann Appl Biol 172:1–3 Saunders ME, Luck GW, Mayfield MM (2013) Almond orchards with
Lichtenberg EM, Kennedy CM, Kremen C, Batáry P, Berendse F, living ground cover host more wild insect pollinators. J Insect
Bommarco R, Crowder DW (2017) A global synthesis of the Conserv 17:1011–1025
effects of diversified farming systems on arthropod diversity Shackelford G, Steward PR, Benton TG, Kunin WE, Potts SG, Bies-
within fields and across agricultural landscapes. Glob Change meijer JC, Sait SM (2013) Comparison of pollinators and natural
Biol 23:4946–4957 enemies: a meta-analysis of landscape and local effects on abun-
McGill BJ, Etienne RS, Gray JS, Alonso D, Anderson MJ, Benecha dance and richness in crops. Biol Rev 88:1002–1021
HK, Dornelas M, Enquist BJ, Green JL, He F, Hurlbert AH, Simon S, Bouvier J-C, Debras J-F, Sauphanor B (2010) Biodiversity
Maguarran AE, Marquet PA, Maurer BA, Ostling A, Soykan CU, and pest management in orchard systems. A review. Agron Sustain
Ugland KI, White EP (2007) Species abundance distributions: Dev 30:139–152
moving beyond single prediction theories to integration within Sisterson MS, Yacoub R, Montez G, Grafton-Cardwell EE, Groves
an ecological framework. Ecol Lett 10:955–1015 RL (2008) Distribution and management of citrus in California:
Moreira X, Abdala-Roberts L, Rasmann S, Castagneyrol B, Mooney implications for management of glassy-winged sharpshooter. J
KA (2016) Plant diversity effects on insect herbivores and their Econ Entomol 101:1041–1050
natural enemies: current thinking, recent findings, and future Sisterson MS, Thammiraju SR, Lynn-Patterson K, Groves RL, Daane
directions. Curr Opin Insect Sci 14:1–7 KM (2010) Epidemiology of diseases caused by Xylella fastidi-
Moulton JK, Wiegmann BM (2007) The phylogenetic relationships osa in California: evaluation of alfalfa as a source of vectors and
of flies in the superfamily Empidoidea (Insects: Diptera). Mol inocula. Plant Dis 94:827–834
Phylogenet Evol 43:701–713 Soares JBC, Costa EM, da Silva FEL, Araujo EL (2016) Insect diver-
Pfister SC, Sutter L, Albrecht M, Marini S, Schirmel J, Entling MH sity in an experimental vineyard in the state of Rio grande do
(2017) Positive effects of local and landscape features on predator norte, Brazil. Revista Caatinga 29:239–245
flies in European agricultural landscapes. Agric Ecosyst Environ Ssymank A, Kearns CA, Paper T, Thompson FC (2008) Pollinating
239:283–292 flies (Diptera): a major contributor to plant diversity and agricul-
Pocock MJO, Jennings N (2008) Testing biotic indicator taxa: the sen- tural production. Trop Conserv 9:86–89
sitivity of insectivorous mammals and their prey to the intensifica- Summers CG (1998) Integrated pest management in forage alfalfa.
tion of lowland agriculture. J Appl Ecol 45:151–160 Integr Pest Manag Rev 3:127–154
Ponce C, Bravo C, de León DG, Magaña M, Alonso JC (2011) Effects Triplehorn CA, Johnson NF (2005) Borror and Delong’s introduction
of organic farming on plant and arthropod communities: a case to the study of insects. Brooks/Cole, Belmont
study in Mediterranean dryland cereal. Agric Ecosyst Environ Wardhaugh CW (2015) How many species of arthropods visit flowers?
141:193–201 Arthropod Plant Interact 9:547–565
Poveda K, Martinez E, Kersh-Becker MF, Bonilla MA, Tscharntke T Winter S, Bauer T, Straus P, Kratschmer S, Paredes D, Popescu D,
(2012) Landscape simplification and altitude affect biodiversity, Landa B, Guzmán G, Gómez JA, Guernion M, Zaller JG, Batáry P
herbivory and Andean potato yield. J Appl Ecol 49:513–522 (2018) Effects of vegetation management intensity on biodiversity
Prischmann DA, James DG, Gingras SN, Snyder WE (2005) Diversity and ecosystem services in vineyards: a meta-analysis. J Appl Ecol
and abundance of insects and spiders on managed and unmanaged 55:2484–2495
grapevines in southcentral Washington state. Pan-Pacific Entomol Yaghmour M, Haviland DR, Fichtner EJ, Sanden BL, Viveros M,
81:131–144 Sumner DA, Stewart DE, Gutierrez CA (2016) Sample costs to
Rader R, Bartomeus I, Garibaldi LA, Garratt MPD, Howlett BG, establish an orchard and produce almonds: San Joaquin Valley
Winfree R, Cunningham SA, Woyciechowski M (2016) Non-bee South. University of California Cooperative Extension, Oakland
insects are important contributors to global crop pollination. Proc
Natl Acad Sci USA 113:146–151 Publisher’s Note Springer Nature remains neutral with regard to
Rieux R, Simon S, Defrance H (1999) Role of hedgerows and ground jurisdictional claims in published maps and institutional affiliations.
cover management on arthropod populations in pear orchards.
Agric Ecosyst Environ 73:119–127
Ruano F, Lozano C, Garcia P, Peña A, Tinaut A, Pascual F, Campos M
(2004) Use of arthropods for the evaluation of the olive-orchards
management regimes. Agric For Entomol 6:111–120

13

You might also like