Food Research International 190 (2024) 114574

Contents lists available at ScienceDirect

Food Research International

journal homepage: www.elsevier.com/locate/foodres

Development of a lexicon for the sensory description of edible insects

commercially available in Australia
Ishka Bless a, b, Susan Elaine Putnam Bastian a, Joanne Gould b, Qian Yang b, Kerry
Leigh Wilkinson a, *
a
School of Agriculture, Food and Wine, The University of Adelaide, Waite Campus, PMB 1, Glen Osmond, 5064 South Australia, Australia
b
School of Biosciences, University of Nottingham, Sutton Bonington Campus, Sutton Bonington, Loughborough LE12 5RD, UK

A R T I C L E I N F O A B S T R A C T

Keywords: Sensory lexicons provide an important tool for describing the sensory properties of emerging, unfamiliar foods
Entomophagy such as edible insects. This study sought to establish and validate a sensory lexicon for the description and
Insects as food differentiation of edible insects commercially available in Australia and prepared using common preservation
Sensory analysis
and cooking methods (freeze-drying, hot-air drying, roasting, sautéing and deep-frying). Five species were
House cricket
evaluated, including house crickets (Acheta domesticus), yellow mealworm larvae (Tenebrio molitor), king meal­
Yellow mealworm
King mealworm worm larvae (Zophobas morio), tyrant ants (Iridomyrmex spp.) and green tree ants (Oecophylla smaragdina).
Green tree ant Following generic descriptive sensory analysis methods, a trained panel (n=8) developed a sensory lexicon of 29
Tyrant ant aroma and flavour descriptors, and 16 texture descriptors. Vocabulary were then categorised and ordered to
generate a sensory wheel. Due to a lack of cross-over in sensory attributes between species, sub-categories of
species-specific vocabulary were also generated for each insect. The lexicon enabled sensory profiling of
commercially available edible insect samples which revealed large variation in aroma, flavour, and texture at­
tributes due to both species and preparation method. This work provides a platform for development of a globally
relevant edible insect sensory lexicon. International collaboration will enable expansion of the lexicon for use
with other insect species and preparation methods, insect-derived ingredients (such as insect powder, defatted
insect powder and textured insect protein) and in different cultural settings. As the industry grows, the appli­
cability of vocabulary for differentiating within species and between competitive products should also be
assessed.

1. Introduction typically lack exposure to insects in the context of food. Rather, insects
are often associated with negative connotations (i.e., as a pest, nuisance
Insects have been identified as an environmentally sustainable and/or source of contamination) (Deroy, Reade, & Spence, 2015; van
source of high-quality protein. Entomophagy (the human consumption Huis, 2013). This is reflected in Western consumer responses to ento­
of insects) therefore offers a promising solution for protein diversifica­ mophagy whereby feelings of disgust are dominant (Bae & Choi, 2021;
tion in Western food cultures (van Huis et al., 2013). Edible-insect- Jensen & Lieberoth, 2019), and considered a primary predictor of con­
related businesses have consequently emerged in regions such as sumer acceptance (La Barbera, Verneau, Amato, & Grunert, 2018).
Europe, North America, Australia, and New Zealand. However, poor Improving consumer familiarity with insects as food could therefore
consumer acceptance remains a major and critical barrier to industry help to redefine entomophagy in Western food cultures (Yang & Lee,
growth (Dagevos, 2021). 2019).
While entomophagy played an important role in early human diets Describing the sensory properties (aroma, flavour, texture, and
(Lesnik, 2018) and continues to be practiced in regions such as Latin appearance) of food products (e.g., through various forms of marketing
America, Africa, and Asia (Raheem et al., 2019), Western consumers such as product labelling and advertisements) can help consumers

Abbreviations: AD, Acheta domesticus; TM, Tenebrio molitor; ZM, Zophobas morio; IR, Iridomyrmex spp; OS, Oecophylla smaragdina; ANOVA, analysis of variance; a,
aroma attribute; f, flavour attribute; m, mouthfeel attribute; t, texture attribute.
* Corresponding author.
E-mail address: [email protected] (K.L. Wilkinson).

https://doi.org/10.1016/j.foodres.2024.114574
Received 22 February 2024; Received in revised form 26 May 2024; Accepted 26 May 2024
Available online 27 May 2024
0963-9969/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
I. Bless et al. Food Research International 190 (2024) 114574

increase familiarity with new and unique product categories (Tuorila & terms (Drake & Civille, 2003). Once prepared, the sensory lexicon was
Hartmann, 2020). This may play a role in improving consumer accep­ used to establish sensory profiles for a selection of edible insect samples
tance by better informing product expectations and culinary use among (described in Section 2.3). This final step was employed to validate the
consumers and other end users (e.g., chefs and product developers) vocabulary before ordering the terms into a tiered system for presen­
(Piqueras-Fiszman & Spence, 2015; Tan et al., 2015; Wassmann, Sieg­ tation as a sensory wheel (Lawless & Civille, 2013).
rist, & Hartmann, 2021). The accurate description of sensory properties Typically, generic descriptive sensory analysis is conducted by an
is also important for various aspects of food production (e.g., product experienced sensory panel, but a new panel may also be recruited and
development and quality control), as well as sensory and consumer trained following various published standards (e.g., ISO 8586:2023)
research (Suwonsichon, 2019). In turn, this enables the commerciali­ (Lawless & Civille, 2013). As a sensory panel familiar with edible insects
sation of high-quality food products which better meet consumer tastes was not available, a new panel was recruited and trained in accordance
and preferences. Sensory language (the vocabulary used to describe with ISO Standards (ISO, 2012). Methods used for sensory panel
sensory properties) can therefore offer a crucial tool for development of recruitment, lexicon development and panel training as well as sensory
the edible insect industry. profiling and lexicon validation are detailed in Sections 2.2 – 2.7. Ethical
Presently, informal sensory language is routinely used by consumers, approval for the involvement of human subjects in this study was
industry, and researchers to describe edible insects. While recent work granted by the University of Adelaide’s Human Research Ethics Com­
has shown that significant variation in the aroma, flavour and texture of mittee, approval reference number H-2021-207, 16/12/2021.
edible insects is attributable to preparation method (Mishyna, Chen, &
Benjamin, 2020; Perez-Santaescolastica, De Winne, Devaere, & Fraeye, 2.2. Sensory panel
2022), this language is often developed at a species-level and lacks in­
formation regarding how the insect was prepared (Elhassan, Wendin, A nine-person sensory panel (eight women and one man, aged be­
Olsson, & Langton, 2019; Nordic Food Lab, Evans, Flore, & Frøst, 2017; tween 25 and 74 years), participated in this study. Panellists were
Ramos-Elorduy & Menzel, 1998). Further, despite playing a critical role selected on the basis of availability (e.g., the panellist was able to attend
in consumer acceptance (Mishyna, Chen, & Benjamin, 2020), texture is all sessions) and motivation (e.g., the panellist showed demonstrated
poorly described. The vocabulary selected for sensory and consumer interest in sensory research, sustainable food and/or edible insects),
research also varies between studies, limiting their comparison (Nervo, then screened for their ability to identify and describe aroma, flavour,
Ricci & Torri, 2024; Sick, Hartmann & Frøst, 2024). Further research is and texture attributes using odour and basic taste identification assess­
therefore needed to establish well-defined vocabulary that can be used ments, as well as flavour detection and intensity tests, and a combined
to more accurately and uniformly describe and differentiate edible in­ description test (ISO, 2012).
sects prepared using common preservation and cooking methods. Following lexicon development and panel training (described in
A sensory lexicon comprises standardised vocabulary for the sensory Section 2.6 and 2.7), one panellist was excluded due to poor reproduc­
attributes of a given food product category. Like a dictionary, it provides ibility (e.g., their inability to consistently score samples relative to other
a list of descriptors (vocabulary used to describe sensory attributes) panel members) (Rossi, 2001). An eight-person sensory panel therefore
paired with a definition and reference standard. The definition clarifies participated in later sensory profiling.
the meaning of the descriptor while the reference standard (a food or
chemical product) provides a benchmark against which the attribute can 2.3. Sample selection
be measured. Thus, facilitating clear and consistent description across
research, industry and among consumers, as well as in different coun­ Samples selected for lexicon development and sensory profiling are
tries and cultures (Suwonsichon, 2019). Previous research has estab­ described in Table 1. For the purpose of lexicon development, sample
lished standardised vocabulary for a small number of edible insect selection was conducted with the aim of providing comprehensive in­
samples. Albrektsson (2017) developed standardised aroma, flavour, sights into potential aroma, flavour and texture attributes within the
and texture vocabulary for edible insects commercially available in product category (Lawless & Heymann, 2010). A sub-set of samples was
Sweden, but common cooking techniques (e.g., roasting, sautéing, and then selected for sensory profiling to validate the developed lexicon.
deep-frying) were not evaluated. Kiatbenjakul, Intarapichet & Cadwal­ Species were chosen on the basis of being available for consumer pur­
lader (2015) have also developed aroma vocabulary for the male giant chase in Australia at the time of this study. These include house crickets
water bug (Lethocerus indicus). This study employed generic descriptive (Acheta domesticus; AD), yellow mealworm larvae (Tenebrio molitor; TM),
sensory analysis methods (Lawless & Heymann, 2010) to establish and king mealworm larvae (Zophobas morio; ZM), tyrant ants (Iridomyrmex
validate a sensory lexicon for the description and differentiation of spp.; IR) and green tree ants (Oecophylla smaragdina; OS). Samples were
edible insects commercially available in Australia − including house purchased from commercial Australian suppliers − Grubs Up (Pinjarra,
crickets (Acheta domesticus; AD), yellow mealworm larvae (Tenebrio Australia), Circle Harvest (Sydney, Australia), and Something Wild
molitor; TM), king mealworm larvae (Zophobas morio; ZM), tyrant ants (Adelaide, Australia) − and prepared using methods of commercial and
(Iridomyrmex spp.; IR) and green tree ants (Oecophylla smaragdina; OS) − traditional relevance (Mishyna et al., 2020). At the time of this study, OS
and prepared using common preservation and cooking methods. were sold and consumed in Australia in their raw form, while IR were
Through collaborative efforts, it is anticipated that the lexicon can sold and consumed as a hot-air dried product. AD, TM and ZM were
provide a starting point for development of a globally relevant commercially available as hot-air dried products only, but frozen sam­
descriptive tool. ples were also sourced and subsequently freeze-dried, roasted, sautéed,
and deep-fried to encompass preservation and cooking methods used in
2. Materials and method other regions (Mishyna et al., 2020). Evaluations were completed using
whole insects only (excluding aroma evaluation, for which freeze-dried
2.1. Experimental design and hot-air dried samples were coarsely ground due to low aroma in­
tensity and sample availability). Insect-derived ingredients (such as in­
This study followed generic descriptive sensory analysis methods sect powder, defatted insect powder, textured insect protein and oils)
described by Lawless and Heymann (2010), which involve four key were excluded from this study due to their lack of availability (only AD
stages: 1) establishing a ‘frame of reference’ by sampling a range of powder was available at the time of the study).
products from a given category, 2) generating terms to describe sensory
attributes of the products, 3) refining the list of terms by consensus, and
4) developing a definition and reference standard for the final list of

2
 I. Bless et al. Food Research International 190 (2024) 114574

2.4. Sample preparation

2 whole insects, 1

insects, 0.1 g per

15 whole ants,
Sample preparation methods for lexicon development and sensory
Flavour and

0.5 g whole
evaluation.

evaluation.

evaluation.
insect per profiling are described in Table 1. Hot-air dried TM and IR were pre­

1 ant per
pared by the supplier and stored in vacuum-sealed bags at 3 ◦ C until
texture

evaluation. All other samples were supplied frozen and stored at – 20 ◦ C

until preparation for evaluation. Freeze-dried samples were prepared by
a commercial freeze-drying facility, then stored in an airtight food
Freeze-dried and hot-air dried samples:

15 whole ants. 4 ants per evaluation,

storage bag at 3 ◦ C. Hot-air dried AD and ZM were prepared the day
sautéed, and deep-fried samples: 2

freshly pressed with filter paper.

0.5 g, coarsely ground. Roasted,

before evaluation, and after cooling to room temperature (21 ± 1 ◦ C)

were stored in vacuum-sealed bags at 3 ◦ C. Roasted, sautéed, and deep-
fried samples were freshly prepared immediately prior to evaluation. OS
Presentation method

0.5 g whole insects. were defrosted at room temperature (21 ± 1 ◦ C) for 1-min prior to
serving.
whole insects.

2.5. Sample presentation

Aroma

Sample presentation formats are described in Table 1 and were

developed separately for each species and preparation method. For in­
75 s at 200 ◦ C in canola oil.

45 s at 200 ◦ C in canola oil.

90 s at 200 ◦ C in canola oil.

300 mL oil / 100 g. Dried

300 mL oil / 100 g. Dried

300 mL oil / 100 g. Dried

dividual evaluations during lexicon development, samples were con­

on paper towel before

on paper towel before

on paper towel before

tained in a small, white plastic bowl labelled with a three-digit code and
presented on a white plastic tray, alongside a plastic spoon and glass of
filtered water. To reflect consumer use, roasted samples were initially
e
Deep-fried

served immediately after preparation. As this method was associated

serving.

serving.

serving.

with considerable temperature variability during evaluation (i.e., sam­

ples quickly cooled to room temperature), samples were re-evaluated at
nil

room temperature (21 ± 1 ◦ C) during lexicon development and sensory

canola oil. 6.25 mL

canola oil. 6.25 mL

canola oil. 6.25 mL

4 min at 175 ◦ C in

2 min at 175 ◦ C in

4 min at 175 ◦ C in

profiling sessions.
For sensory profiling, samples were presented in lidded, plastic petri
Samples roasted at two temperatures for lexicon development, 160 ◦ C and 180 ◦ C; and 160 ◦ C only for sensory profiling.
oil / 100 g.

oil / 100 g.

oil / 100 g.

dishes labelled with three-digit codes. To aid temperature control of hot

e
Sautéed

samples (e.g., sautéed and deep-fried) during the longer evaluation,

petri dishes were lined with aluminium foil. All samples were presented
nil

on a white plastic tray, alongside a plastic spoon, a glass of filtered water

and fresh green apple slices (for palate cleansing). A piece of filter paper
min at 180 ◦ C.d

min at 180 ◦ C.d

min at 180 ◦ C.d

160 ◦ C or 15

160 ◦ C or 10

160 ◦ C or 20

was also provided for aroma evaluation of OS (allowing panellists to

20 min at

12 min at

25 min at

burst the ant abdomen prior to evaluation).

e
Roasted

nil

2.6. Development of a sensory lexicon

moisture < 5 %.

moisture < 5 %.

Panellists initially received general training in the detection and

e
Hot-air dried

Commercial

Commercial
75 ◦ C until

75 ◦ C until

description of aroma, flavour, and texture attributes across 3 × 3 hr

Sample preparation and presentation for lexicon development and sensory profiling.

productb

productb

sessions. They were then presented with each sample over 2 × 2 hr

sessions, and asked to individually evaluate the aroma, flavour, and
texture – describing any attributes that they identified, without the use
Preparation method

Other (served fresh).

of hedonic or quantitative terms. For their initial evaluation, panellists

mBar, condenser

were required to follow a specific assessment technique for aroma,

4 days, 0.5–1
Freeze-dried

flavour, and texture evaluation (standardised to normal consumption).

at − 30 ◦ C.

Sample purchased from Something Wild (Adelaide, Australia).

They were then invited to use alternative assessment techniques if

Samples purchased from Circle Harvest (Sydney, Australia).

desired but asked to record their method. This allowed for the optimi­
Samples blanched for 5 min at 100 ◦ C before preparation.
nil

Samples purchased from Grubs Up (Pinjarra, Australia).

sation of assessment techniques used for sensory profiling sessions, as

described in Supplementary Materials, Section A. Descriptors were
Acheta domesticus (AD)

Oecophylla smaragdina

subsequently collated and presented to the panel alongside the samples

Iridomyrmex spp. (IR)
Tenebrio molitor (TM)

Zophobas morio (ZM)

and a range of potential reference standards for each descriptor (Lawless

Scientific name

& Heymann, 2010). The consensus method was then followed to refine
the list of descriptors and remove redundancy, as well as develop
descriptor definitions and final reference standards.
(OS)

2.7. Sensory profiling and lexicon validation

Yellow mealworm larvaeb

King mealworm larvaea

2.7.1. Panel training

Following lexicon development, the panel was trained in sensory
profiling of edible insect samples. Training sessions were conducted over
Green tree antsc
House cricketsa

4 × 3 hr sessions. In accordance with ISO Standards (ISO, 2012), pan­

Tyrant antsb

ellists were trained in the use of a 10 cm unstructured line scale

Sample
Table 1

(anchored ‘low’ at 1 cm and ‘high’ at 9 cm) for aroma, flavour and

texture evaluation – progressing from simple (e.g., single attribute,
d
b
a

e
c

3
I. Bless et al. Food Research International 190 (2024) 114574

Table 2
Aroma and flavour descriptors, definitions, and reference standards for house crickets (Acheta domesticus; AD), yellow mealworm larvae (Tenebrio molitor; TM), king
mealworm larvae (Zophobas morio; ZM), tyrant ants (Iridomyrmex spp.; IR) and green tree ants (Oecophylla smaragdina; OS).
Descriptor Definition Reference standardb c
Insect
speciesd

Cereal Aroma and flavour of cereals, such as barley, oat 30 g equal ratio of McKenzie’s Raw Barley, Black and Gold Rolled Oats and coarsely AD, TM, ZM
and wheat. ground Weet-bix; mixed and submerged in distilled water.
Toasted Aroma and flavour of toasted bread crust. 2 tbsp Tip Top ‘The One’ White Sandwich Sliced Bread, toasted until golden brown on AD, TM, ZM
each side. Cooled, then coarsely ground.
Popcorn* Aroma and flavour of freshly popped popcorn. 2 tbsp Riviana Popping Corn, popped in 2 tbsp vegetable oil per ¼ cup corn kernels. AD, TM, ZM
Popcorn popped within 24 h of use and coarsely ground.
Oily Aroma and flavour of oil. 2 tbsp Black and Gold Vegetable Oil. AD, TM, ZM
Hot chips (fries) Aroma and flavour of frying oil, similar to hot 30 g McCain Superfries (straight cut), deep-fried until crispy in vegetable oil. Cooled AD, TM, ZM
chips. Oily, caramelised and starchy notes. to room temperature and chopped into 2 cm pieces.
Raw walnut Aroma and flavour of raw walnut; not freshly 30 g Lucky natural golden walnuts, coarsely ground. TM
deshelled, producing slight rancidity.
Roasted nuts Aroma and flavour of roasted, mixed nuts. 30 g mixed nuts (equal ratio Coles roasted, unsalted peanuts; Lucky toasted AD, TM, ZM
pistachios; Lucky oven roasted almonds and Lucky natural golden walnuts; coarsely
ground).
Nutritional yeast Aroma and flavour of nutritional yeast; cheesy 2 tbsp Premium Choice Nutritional Savoury Yeast Flakes. TM, ZM
and yeasty notes.
Roasted chicken Aroma and flavour of roasted chicken skin; 30 g skin of a Foodland Roasted Chicken (leg cut). AD, TM, ZM
caramelised and meaty notes.
Bacon* Aroma and flavour of fried bacon. 30 g Primo Pan-sized, Triple Smoked Rindless Bacon, fried until golden brown and ZM
crispy on each side; finely diced.
Beef stock Aroma and flavour of rich, savoury stock; beefy 1 crumbled Massel’s Beef Style Stock Cube and 0.5 g Vegemite, mixed. IR, AD
and soy sauce notes.
Earthy Aroma and flavour of raw mushroom and wet 2 tbsp equal ratio coarsely chopped button mushrooms and wet soil. AD, TM, ZM
soil; similar to a damp forest floor.
Wet haya Aroma of wet hay. 2 tbsp coarsely ground Garden Essentials Sugarcane Mulch, moistened with water. TM
Warm spices Aroma and flavour of warm spices such as 1 tbsp 10:2 ratio, MasterFoods Allspice (ground) and MasterFoods Peppercorns Black IR
cinnamon, clove and pepper. (cracked).
Woody Aroma and flavour of toasted wood; dry and 2 tbsp American Oak, small grade-screened (heavy toast). IR
earthy notes.
Burnt match The aroma and flavour of a burnt match; burnt 4 freshly struck matches, burnt approx. halfway down match. IR
wood and slightly sulphurous.
Rubberf The flavour of rubber; tarry notes. 30 g 4:1 ratio of Bic Glue Stick glue and Marbig Rubber Band Assorted (cut into 1 cm ZM
pieces).
Boiled egg yolk Aroma and flavour of cooked egg yolk. Egg yolk from egg boiled for 6 min. AD, TM, ZM
Seaweed An oceanic aroma and flavour; seaweed and salty 30 g coarsely chopped dried kelp. AD, ZM
notes.
Dried shrimp The aroma and flavour of dried shrimp. 30 g coarsely chopped Richmond dried shrimp. AD, ZM
Burnt toast The aroma and flavour of burnt toast; cereal and 2 tbsp Tip Top ‘The One’ White Sandwich Sliced Bread, toasted until a black surface IR
chemical. on each side is reached. Cooled, then coarsely ground.
Vinegar The sharp, acidic aroma and flavour of vinegar. 2 tbsp Cornwell White Vinegar. OS, IR
Green apple Aroma and flavour of freshly cut green apple with 30 g finely diced granny smith apple, 5 g lemon rind, 5 g lime rind, ½ tsp fresh lemon OS
a hint of citrus. juice.
Fruit candy Aroma and flavour of fruit candy. 2 tbsp finely ground tropical Skittles, equal parts of each flavour. OS
Minta Cooling, minty, menthol aroma. 2 tbsp 6:1 ratio Fisherman’s Friend Original Mint (finely ground) and Hoyt’s dried OS
mint leaves.
Cut grassa The aroma of freshly cut grass; green, fresh. 30 g freshly cut grass, moistened with distilled water. OS
Lingering flavourf Presence of a lingering flavour 20 s after nil OS, IR, AD,
swallowing (or expectoration). TM, ZM
Saltyf Salty taste. 50 mL sodium chloride (5 g/L water). AD, TM, ZM
Savouryf Savoury taste (umami). 50 mL monosodium glutamate (0.6 g/L water). IR, AD, TM,
ZM
Sourf Sour taste. 50 mL citric acid (1 g/L water). OS, IR
Bitterf Bitter taste. 50 mL caffeine (0.5 g/L water). IR, AD, TM,
ZM
a
Aroma descriptor only.
f
Flavour descriptor only.
b
All reference standards are considered equivalent to nine (90 = high) on a scale of 0–100.
c
All aroma and flavour reference standards were presented in 3.8 cm × 5 cm clear lidded containers for sensory profiling sessions.
d
This descriptor was included in the species-specific lexicon for the listed insect species.
* Descriptor removed from final lexicon (as described in Section 3.2.4).

water-based sample) to complex evaluations (e.g. multiple attributes, selection was completed as per ISO Standards (ISO, 2012).
food sample). Panellists then selected assessment techniques, a palate
cleanser (fresh green apple slices) and descriptor list for each species by 2.7.2. Sensory profiling
consensus, before completing product specific training with edible in­ Sensory profiling sessions were completed in a 12-booth sensory
sect samples (reflecting final evaluation conditions and requirements). laboratory at the University of Adelaide’s Waite Campus, under
Due to the time-sensitive nature of textural properties during sensory controlled conditions (21 ± 1 ◦ C, pure white LED lighting). Each booth
profiling, reference products for texture were only provided during was equipped with a computer and responses were recorded using
product training – allowing panellists to become familiar with each RedJade® software (Redwood City, CA, USA). Samples were presented
attribute and reach consensus for low and high intensity. Final panel (in triplicate) using a monadic sequence and randomised block design.

4
I. Bless et al. Food Research International 190 (2024) 114574

Table 3
Texture descriptors, definitions, and reference standards for house crickets (Acheta domesticus; AD), yellow mealworm larvae (Tenebrio molitor; TM), king mealworm
larvae (Zophobas morio; ZM), tyrant ants (Iridomyrmex spp.; IR) and green tree ants (Oecophylla smaragdina; OS).
Descriptor Definitiona Reference standard Insect speciesb

Low intensity High intensity

Burst-in- The sensation of a sudden popping upon bite-down, releasing 1 × 1 cm cube of fresh orange. 1 × Chatime Bubble Tea OS
moutht interior contents. Popping Pearl.
Pricklyt Sharp protruding points. nil nil IR, AD, TM, ZM
Brittlet Shattering fracture (many pieces), following a relatively small 1 × Captain’s Table water 2 × 2 cm piece Cruskit IR, AD, TM, ZM
amount of force. cracker. Original cracker.
Tought Malleable and resistant to fracture; bending/compressing in 2 × 2 cm piece Coles Bakery 2 cm piece Coles dried OS, AD, TM, ZM
response to biting force. White Sourdough, crusts Mango Slices.
removed.
Denset A measure of air-pockets incorporated into food structure, where 2 × 2 cm piece Cruskit Original 1 × 1 cm piece Coles roasted AD, TM, ZM
low refers to the presence of large air pockets, medium refers to the cracker. chicken (breast, no skin).
presence of small air pockets and high refers to no air pockets.
Crunchyt A textured food which fractures during mastication and produces 1 × 1 cm piece honeycomb. 2 × 2 cm piece Cruskit IR, AD, TM, ZM
relatively loud ‘crackly’ sounds. Original cracker.
Chewyt A food that does not break into pieces during mastication and 2 × 2 cm cube Coles Bakery 1 × The Natural OS, AD, TM, ZM
requires drawn-out chewing to prepare for swallowing. White Sourdough, crusts Confectionary Co. Snake.
removed.
Crumblyt A food that breaks into small, crumb-like pieces during mastication. 2 cm piece Arnott’s Scotch 1 × Always Fresh Mini Toasts. AD, TM, ZM
The food does not require drawn-out chewing to prepare for Finger biscuit.
swallowing.
Flakyt A food that breaks into coarse, flake-like pieces during mastication. 2 × 2 cm piece Cruskit Original 2 cm piece Cadbury Flake. IR, AD, TM, ZM
The food does not require drawn-out chewing to prepare for cracker.
swallowing.
Particulatem The sensation of particulates in the mouth during mastication. Can ¼ tsp cornflour (low); ¼ tsp ¼ tsp Macro Australian IR, AD, TM, ZM
be analysed separately as low/chalky, medium/grainy, high/gritty. Macro Natural Semolina Flour Polenta.
(medium).
Wads-upm The aggregation of the food during mastication, forming a pasty 1 × Captain’s Table water 2 × 2 cm cube Coles Bakery AD, TM, ZM
‘wad’ in the mouth. cracker. White Sourdough, crusts
removed.
Hard-to-clearm The sensation of small pieces catching in- or sticking to the teeth or nil nil OS, IR, AD, TM, ZM
throat while or after swallowing.
Dryingm A drying sensation on the tongue and in the mouth after swallowing. 1 × Captain’s Table water 2 cm piece Cruskit Original IR, AD, TM, ZM
cracker. cracker.
m
Astringent Puckering sensation, primarily on the surface of the tongue and 50 mL tannic acid (0.05 g/L 50 mL tannic acid (1.5 g/L OS, IR
walls of mouth. Can leave the mouth feeling rough, raw, or water). water).
sandpapery.
Mouthcoatingm The sensation of having a coating across mouth surfaces. Vegetable oil Lard AD, TM, ZM
Tooth The sensation of food building-up in- and sticking on the surface of 1 × Captain’s Table water 1 × The Natural AD, TM, ZM
packingm teeth, particularly in molars. cracker. Confectionary Co. Snake.
t
Texture descriptor.
m
Mouthfeel descriptor.
a
Specific description for low and high intensity are provided in Supplementary Materials, Section A.
b
This descriptor was included in the species-specific lexicon for the listed insect species.

For each sample, panellists were asked to evaluate the intensity of 2.7.4. Statistical analysis
selected aroma, flavour, and texture attributes on an unstructured line Statistical analysis was completed using XLSTAT software Version
scale, anchored from low to high (if the attribute was not perceptible, 2023.1.5 (Addinsoft, Paris, France). To assess the importance of each
the scale was left blank). Evaluations followed a standardised sample descriptor for describing and differentiating sensory attributes, a three-
assessment technique, described in Supplementary Materials, Section A. way mixed model analysis of variance (ANOVA) was fitted for AD, TM
Panellists were also provided with a set of reference aroma and flavour and ZM samples, with the assessor as a random effect, and replicate and
standards, as well as a copy of the lexicon (shown in Supplementary product as fixed effects. A post-hoc Tukey’s honestly significant differ­
Materials, Section A). Sampling was conducted across 6 sessions, with 2 ence (HSD) test was then applied to identify significant differences be­
sessions per day (8–9 samples per session) and a 15 min break between tween samples prepared using different preservation and cooking
same-day sessions. A 1 min break was also provided between samples, methods. As AD, TM and ZM samples were evaluated using species-
during which panellists were required to palate cleanse using filtered specific sub-lexicons, a separate ANOVA was fitted for each species to
water and fresh green apple slices. This protocol was designed in prevent zero-inflation. Multiple factor analysis (MFA) was then con­
conjunction with panellists to manage fatigue and maximise their ducted on differentiating sensory attributes (P ≤ 0.05) across AD, TM
availability, and evaluated by examining the replicability of panellist and ZM to visualise the relationship between preparation method and
scores during training. sensory profiles, and explore trends between different species. In the
absence of comparative products (only one sample was available for
2.7.3. Lexicon validation each species), mean attribute intensity scores were calculated to eval­
Sensory profiles for each insect species were used to validate the uate sensory language for IR and OS.
vocabulary (e.g., confirming whether the terms enable description and
differentiation of the product category) (Drake & Civille, 2003), as
described in Section 2.7.4. The final list of descriptors was then cat­
egorised and ordered to generate a sensory wheel.

5
I. Bless et al. Food Research International 190 (2024) 114574

3. Results and discussion

3.1. Preliminary sensory lexicon development

disagreement relating to these and other attributes (such as dried shrimp)
A preliminary list of 47 terms was generated during lexicon devel­ was also observed during lexicon development. However, following
opment, comprising 31 aroma and flavour descriptors, and 16 texture sample re-evaluation and additional panel training, all attributes were
descriptors (including texture and mouthfeel attributes). All terms as retained in the preliminary sensory lexicon because the panel perceived
well as their definition and reference standard are detailed in Tables 2 that they were important to product use and acceptability. Continued
and 3. Like the ‘Lexicon for Australian Native Plant Foods and Ingredients’ differences in evaluation may therefore reflect long-term, deep-rooted
described by Smyth et al. (2012), the broad product category resulted in influences such as product familiarity (Jeong & Lee, 2021). While sen­
a lack of cross-over between attributes – in this case, between species. sory panel performance was considered satisfactory for the purpose of
For example, AD, TM and ZM shared no aroma and flavour descriptors validating lexicon applicability, future work should explore Rate-All-
with OS. This distinction between insect species was also observed by That-Apply (RATA) and hedonic testing with untrained consumers to
Mishyna et al. (2020). To reduce redundancy and panel fatigue, sub- determine whether a) affected attributes are perceived at a population
categories of species-specific lexicons were therefore generated for level, and b) these attributes influence consumer acceptability and
each insect. Notably, dried shrimp and seaweed were excluded from the preference.
TM lexicon despite previous description in roasted and deep-fried TM While some variability was expected between samples (e.g., due to
(Seo, Kim, & Cho, 2020). As these attributes are associated with the biological variation) (Meyer-Rochow, Gahukar, Ghosh, & Jung, 2021),
presence of pyrazines (2,6-dimethyl-pyrazine, 2,3,5-trimethyl-pyra­ there was no significant (P ≤ 0.05) replicate × assessor effect for most
zine), heptan-2-one and 1-butyl-pyrrolidine, this could be explained attributes. However, a higher number of attributes with a significant
by panel selection of alternative savoury-type descriptors including replicate × assessor effect was observed for AD. Sample variability may
roasted nuts, toasted, nutritional yeast and boiled egg yolk. have therefore had a greater effect on product discrimination for this
Another key challenge for this product category relates to sample species. Alternatively, panellist fatigue may be higher when tasting AD
availability. Typically, a panel would be exposed to a large sample set samples due to more complex textural properties (e.g., large legs and
(frame of reference) of approximately 25–100 products (Drake & Civille, wings), reducing repeatability.
2003). However, given Australia’s relatively small edible insect in­
dustry, a limited number of products were commercially available. The 3.2.2. Describing and differentiating by species
frame of reference used for lexicon development was therefore restricted IR and OS can be described by distinct sensory profiles shown in
to one product sample per insect species, prepared using relevant pres­ Fig. 1. IR was characterised by burnt match, woody and savoury beef stock
ervation and cooking methods. While this reflects the Australian market attributes, a moderately bitter taste, lingering flavour and a brittle, prickly
at the time of this study, samples from other suppliers may exhibit texture, that was grainy upon mastication. After swallowing, the sample
diverse sensory properties (e.g., due to changes in rearing, processing was moderately hard-to-clear, drying and astringent. OS was characterised
and/or preparation methods) (Mishyna et al., 2020). As the industry by green apple, cut grass and vinegar aromas. Upon bite down, the sample
expands, future work should therefore include a more detailed investi­ was mildly tough and burst-in-mouth, releasing a strong sour taste and
gation of sensory variation within insect species and between competi­ green apple and vinegar flavours. This characteristic sourness alongside
tive products. This should incorporate samples from the international fruity aroma and flavour could be explained by the high concentration of
market as well as edible insect species of cultural and commercial sig­ volatile organic acids (predominantly formic acid) in OS (Alagappan
nificance (not currently available for sale in Australia) − providing a et al., 2021). 1-Hexanol has also been reported in smaller quantities in
globally relevant language that incorporates traditional and commercial the anterior of the ant, possibly accounting for the presence of cut grass
preparations. As sautéed and deep-fried samples were prepared using aroma (Alagappan et al., 2021). Upon mastication and swallowing, the
only canola oil, the applicability of this lexicon for evaluating insects sample was moderately chewy and hard-to-clear with an astringent
cooked in other frying oils could also be further explored. mouthfeel. Across both IR and OS samples, secondary (lower intensity)
attributes were also demonstrated (e.g., attributes with mean intensity
3.2. Sensory profiling and lexicon validation scores of < 40). As these qualities may play a role in differentiating
commercial products as well as consumer preference (e.g., dominance of
The preliminary sensory lexicon was evaluated for use in the savoury versus warm spices flavours, and artificial/fruit candy versus
description and differentiation of IR, OS, AD, TM and ZM, by species and fresh/green apple aroma and flavour), their inclusion in the lexicon is
preparation method. As species-specific aroma, flavour and texture at­ warranted at this stage of market and research development. A review of
tributes were used for the evaluation of each insect, sensory profiles are selected species-specific descriptors is recommended when additional
reported separately for IR, OS, AD, TM and ZM. commercial products become available. Where possible, this should
include other preparations such as fresh, freeze-drying and hot-air
3.2.1. Panel performance and sample variability drying.
Panel performance and sample variability (e.g., variation between The sensory profiles for AD, TM and ZM are described in Supple­
replicates) were examined by species, using results from the sensory mentary Materials (Section C) and summarised in Fig. 2. While prepa­
profiling of AD, TM and ZM. Results are summarised in Supplementary ration method significantly altered aroma, flavour, and texture
Materials, Section B. IR and OS were not considered due to the absence attributes, all three species were characterised by savoury qualities,
of comparative samples (only one sample was available for each lingering flavour, and complex textural profiles, reflecting their
species). morphology (e.g., exoskeleton, legs, wings, and antennae). Despite these
Examination of the product × assessor effect across AD, TM and ZM similarities, the species could be separated by the source of savouriness
indicates that panellists were evaluating attributes differently (P ≤ and secondary attributes.
0.05), which is common for descriptive sensory analysis methods TM and ZM samples exhibited nutty savoury characteristics,
(Lawless & Heymann, 2010). The comparison of significant product ef­ described by roasted nuts and nutritional yeast aroma and flavour, while
fect and product × assessor effect F-values (P ≤ 0.05) indicates that a AD showed more meat-like qualities with dominant savoury attributes
small group of differentiating attributes (roasted nuts, earthy, oily, drying including roasted chicken, roasted nuts, and beef stock. All samples shared
and mouthcoating) were most affected, with all showing a larger F-value moderate cereal attributes, and mild saltiness. ZM could be separated by a
for the product × assessor effect than the product effect. Panel distinct rubber flavour as well as secondary seaweed and dried shrimp

6
I. Bless et al. Food Research International 190 (2024) 114574

attributes. These qualities could be explained by a higher concentration been reported previously in AD following freeze-drying and blanching
of phenol and trimethylamine (Perez-Santaescolastica et al., 2022; (Perez-Santaescolastica et al., 2022, 2023). TM could also be distin­
Perez-Santaescolastica, De Winne, Devaere, & Fraeye, 2023). Phenol is guished from ZM and AD by the presence of wet hay and raw walnut
known to contribute tarry notes while trimethylamine is a seaweed qualities which were significantly more intense in hot-air dried samples
odour compound (Khatun et al., 2021; Tzompa-Sosa, Yi, Van Valenberg, (P ≤ 0.05 and P ≤ 0.0001 respectively).
& Lakemond, 2019). Seaweed and dried shrimp attributes were also
present in AD. This may be due to the presence of 1-octen-3-ol, which 3.2.3. Describing and differentiating by preparation method
contributes earthy, seaweed, fatty, mould, and mushroom notes, and has Significant differences by Tukey-Kramer honestly significant differ­
ence are shown in Fig. 2 for AD, TM and ZM samples respectively.
Nineteen attributes were found to be significantly different between TM
samples (7 aroma, 6 flavour and 6 texture), 22 between ZM samples (8
aroma, 7 flavour and 7 texture) and 14 between AD samples (4 aroma, 4
flavour and 6 texture). To explore the respective effects of species and
preparation method, Multiple Factor Analysis (MFA) was employed
using differentiating attributes (P ≤ 0.05) across AD, TM and ZM. The
coordinates of the projected points and MFA biplot are shown in Fig. 3.
The first two factors (F1 and F2) explain 78.80 % of variance be­
tween AD, TM and ZM samples, with F1 (x-axis) explaining over half of
variation in the data (54.71 %) and F2 (y-axis) explaining the remaining
24.09 %. The coordinates of the projected points show sautéed and deep-
fried samples clustered to the right of the plot, while hot-air dried,
freeze-dried, and roasted samples are spread across the left. Further, AD,
TM and ZM demonstrate lesser spread when prepared by sautéing and
deep-frying versus hot-air drying, freeze-drying and roasting. This in­
dicates that sautéing and deep-frying may exert a stronger influence on
the sensory profile of AD, TM and ZM samples versus other preparation
methods. These samples were described by hot chips, oily and roasted
chicken aroma and flavour attributes, as well as chewy and dense texture
attributes, clustered on the right side of the biplot.
Hot-air dried, freeze-dried, and roasted samples were mostly
described by attributes on the left side of the biplot. By interpreting
these results alongside Fig. 2, it is evident that hot-air dried AD, TM and
ZM samples were characterised by brittle, crunchy, crumbly, and drying
texture attributes. However, their differentiating aroma and flavour
attributes varied by species. Raw walnut and wet hay were strongly
associated with hot-air dried TM while seaweed, toasted, and earthy were
associated with hot-air dried ZM. Hot-air dried AD was characterised by
a significant increase in earthy aroma and toasted flavour. Squared cosine
values for freeze-dried samples indicate better explanation across F2
(0.806), with differentiating attributes mostly grouped in the lower
quadrants of the biplot. Collectively, these samples were characterised
by a brittle texture. Freeze-dried TM and ZM were also associated with
nutty savoury attributes (nutritional yeast and roasted nuts) while freeze-
dried AD samples were characterised by broad savoury and cereal notes
of moderate intensity. Squared cosine values for roasted samples (0.778)
indicate better explanation across F3 (not shown), reflecting shared at­
tributes with other preparation methods.
Based on the combined analyses of AD, TM and ZM, it is evident that
preparation method caused large variation in insect aroma, flavour, and
texture. This can be mainly attributed to dry heat transfer cooking
methods which are associated with reactions such as Maillard browning
and water evaporation (Perez-Santaescolastica et al., 2022; van Boekel,
2006). Maillard browning can impart savoury, meaty, roasted, toasted,
and bitter notes while water evaporation can affect texture, making
foods crunchy and brittle. Additionally, methods using oil can alter
aroma and flavour, leading to dominant oily, fatty, and savoury attri­
butes (Chang, Wu, Zhang, Jin, & Wang, 2020). These combined factors
could account for the distinct profiles and clustering of sautéed and
deep-fried samples. In contrast to dry heat transfer cooking methods,
freeze-drying has been found to increase sensitivity to lipid oxidation,
producing higher hexanal and 2-methyl propanal levels in insect sam­
ples (Khatun et al., 2021). This may explain the dominant nutty and
cereal notes among freeze-dried samples as hexanal is associated with
Fig. 1. Radar charts of mean attribute intensity scores for tyrant ants (Irido­ green and fatty aromas while 2-methyl propanal is characterised by
myrmex spp.; IR) and green tree ants (Oecophylla smaragdina; OS). (a) Aroma fatty, malted and nutty aromas.
attributes, (b) Flavour attributes, (c) Texture attributes. Finally, the sensory changes observed in samples that were hot-air

7
I. Bless et al. Food Research International 190 (2024) 114574

Fig. 2. Heatmap representing mean attribute intensity scores (rated from 0 to 100) for house cricket (Acheta domesticus; AD), yellow mealworm larvae (Tenebrio
molitor; TM) and king mealworm larvae (Zophobas morio; ZM) samples. FD = Freeze-dried, HD = Hot-air dried, RD = Roasted, SD = Sautéed, DF = Deep-fried.
Different letters (A, B, C) denote a statistically significant difference between samples of the same species for the given attribute (by Tukey-Kramer honestly sig­
nificant difference). Black squares denote attributes that were not evaluated by panellists.

8
I. Bless et al. Food Research International 190 (2024) 114574

Fig. 3. Multiple factor analysis of differentiating attributes (P ≤ 0.05) for house cricket (Acheta domesticus; AD), yellow mealworm larvae (Tenebrio molitor; TM) and
king mealworm larvae (Zophobas morio; ZM) samples (axes F1 and F2: 78.80 %). (a) coordinates of the projected points, (b) biplot. a = aroma attribute, f = flavour.

dried are particularly noteworthy as all three species demonstrated method. However, across all species popcorn and bacon were not char­
significant increases in undesirable attributes such as earthy, seaweed, acterised as differentiating attributes (P ≤ 0.05), nor were they amongst
dried-shrimp and rubber. The distinct raw walnut characteristic in TM was more dominant aromas or flavours. These descriptors were therefore
also perceived as a negative attribute by some panellists. Given this is considered unnecessary and removed from the final sensory lexicon.
the main commercial preparation method in Australia, further research While some authors have identified popcorn and bacon in TM (Wendin,
is warranted to determine to what extent these differences might affect Mårtensson, Djerf, & Langton, 2020; Żołnierczyk & Szumny, 2021), this
consumer liking and preference − providing important insights for early could be related to factors such as biological differences between sam­
product development. ples or cross-cultural effects on sensory perception. Further validation of
the final sensory lexicon with a larger sample range and in other regions
3.2.4. Development of a final sensory lexicon and wheel is therefore warranted.
The results of sensory profiling confirmed applicability of the pre­ The final sensory lexicon, comprising 29 aroma and flavour de­
liminary sensory lexicon for the description and differentiation of edible scriptors and 16 texture descriptors, is described in Tables 2 and 3. A
insects commercially available in Australia by species and preparation sensory wheel was developed by categorising and ordering descriptors

9
I. Bless et al. Food Research International 190 (2024) 114574

to generate a three-tiered wheel, depicted in Fig. 4. The inner tier con­ and needs.
tains sensory properties, including aroma and flavour, as well as texture Overall, this work provides a platform for development of a globally
(comprising texture and mouthfeel attributes). The second tier contains relevant edible insect sensory lexicon. International collaboration will
sub-categories, within which each descriptor was grouped, and the third be required to evaluate and expand existing vocabulary for use with
tier contains all descriptors. To improve visual communication, colour- other insect species and in different cultural settings. This should pri­
coding and images were incorporated for each sub-category. oritise edible insect species of both traditional and commercial signifi­
The vocabulary presented in the sensory lexicon and wheel reflect a cance, appropriate translation into different languages and the selection
combination of previously described and new attributes, therefore of locally available reference standards. As the edible insect industry
expanding-on and standardising current description tools (Albrektsson, grows, ongoing work will also be needed to validate the lexicon for
2017; Mishyna et al., 2020; Nervo, Ricci & Torri, 2024; Sick, Hartmann differentiation within species and between competitive products. This
& Frøst, 2024). should include samples prepared using different rearing, processing,
preservation, and cooking methods, as well as new insect-derived in­
4. Conclusions gredients and foods (such as insect powder, defatted insect powder and
textured insect protein). With this work complete, the sensory lexicon
This study established a sensory lexicon for the description and dif­ and wheel will provide an indispensable tool for clear and consistent
ferentiation of edible insects that are commercially available in description of edible insects across research, industry and among
Australia. Following use of the lexicon for descriptive sensory profiling, consumers.
unnecessary terms (popcorn and bacon) were identified and removed.
Final vocabulary were then categorised and ordered to generate a sen­ Funding
sory wheel, providing a visual communication tool for research and
industry use. Due to the broad nature of the product category, sub- This research was funded by the University of Adelaide School of
categories of species-specific terms were also generated for each insect. Agriculture, Food, and Wine and conducted with support from an
The lexicon enabled sensory profiling of AD, TM, ZM, IR and OS, Australian Government Research Training Program Scholarship.
demonstrating significant variation by both species and preparation
method. This emphasises the broad and versatile nature of edible insects CRediT authorship contribution statement
as a food product category. For example, species and preparation
methods can be selected to produce unique and diverse flavour and Ishka Bless: Writing – original draft, Visualization, Validation,
texture profiles for a wide range of culinary and food product applica­ Methodology, Investigation, Formal analysis, Data curation, Conceptu­
tions. Further, they can be altered to better meet consumer preference alization. Susan Elaine Putnam Bastian: Writing - review & editing,

Fig. 4. Sensory lexicon for edible insects commercially available in Australia.

10
I. Bless et al. Food Research International 190 (2024) 114574

Supervision, Resources, Conceptualization. Joanne Gould: Writing – Food Quality and Preference, 64, 120–125. https://doi.org/10.1016/j.
foodqual.2017.10.002
review & editing, Supervision, Resources, Conceptualization. Qian
Lawless, H. T., & Heymann, H. (2010). Sensory evaluation of food: principles and practices
Yang: Writing – review & editing, Supervision, Resources, Conceptu­ (Vol. 2): Springer.
alization. Kerry Leigh Wilkinson: Writing – review & editing, Super­ Lawless, L. J. R., & Civille, G. V. (2013). Developing lexicons: A review. Journal of Sensory
vision, Resources, Project administration, Funding acquisition, Studies, 28(4), 270–281. https://doi.org/10.1111/joss.12050
Lesnik, J. J. (2018). Edible insects and human evolution. University Press of Florida.
Conceptualization. Meyer-Rochow, V. B., Gahukar, R. T., Ghosh, S., & Jung, C. (2021). Chemical
composition, nutrient quality and acceptability of edible insects are affected by
Declaration of competing interest species, developmental stage, gender, diet, and processing method. Foods, 10(5),
1036. https://doi.org/10.3390/foods10051036
Mishyna, M., Chen, J., & Benjamin, O. (2020). Sensory attributes of edible insects and
The authors declare that they have no known competing financial insect-based foods – Future outlooks for enhancing consumer appeal. Trends in Food
interests or personal relationships that could have appeared to influence Science & Technology, 95, 141–148. https://doi.org/10.1016/j.tifs.2019.11.016
Nervo, C., Ricci, M., & Torri, L. (2024). Understanding consumers attitude towards
the work reported in this paper. insects as food: Influence of insect species on liking, emotions, sensory perception
and food pairing. Food Research International, 182, Article 114174. https://doi.org/
Data availability 10.1016/j.foodres.2024.114174
Nordic Food Lab, Evans, J., Flore, R., & Frøst, M. B. (2017). On eating insects: Essays,
stories and recipes. Phaidon Press.
Data will be made available on request. Perez-Santaescolastica, C., De Winne, A., Devaere, J., & Fraeye, I. (2022). The flavour of
edible insects: A comprehensive review on volatile compounds and their analytical
assessment. Trends in Food Science & Technology, 127, 352–367. https://doi.org/
Acknowledgements
10.1016/j.tifs.2022.07.011
Perez-Santaescolastica, C., De Winne, A., Devaere, J., & Fraeye, I. (2023). Comparing the
The authors gratefully acknowledge Dr Lira Souza Gonzaga for aromatic profile of seven unheated edible insect species. Food Research International,
providing technical support during lexicon development and panel 164, Article 112389. https://doi.org/10.1016/j.foodres.2022.112389
Piqueras-Fiszman, B., & Spence, C. (2015). Sensory expectations based on product-
training, as well as the involvement of sensory panellists that partici­ extrinsic food cues: An interdisciplinary review of the empirical evidence and
pated in this study. theoretical accounts. Food Quality and Preference, 40, 165–179. https://doi.org/
10.1016/j.foodqual.2014.09.013
Raheem, D., Carrascosa, C., Oluwole, O. B., Nieuwland, M., Saraiva, A., Millán, R., &
Appendix A. Supplementary material Raposo, A. (2019). Traditional consumption of and rearing edible insects in Africa,
Asia and Europe. Critical Reviews in Food Science and Nutrition, 59(14), 2169–2188.
Supplementary data to this article can be found online at https://doi. https://doi.org/10.1080/10408398.2018.1440191
Ramos-Elorduy, J., & Menzel, P. (1998). Creepy crawly cuisine: The gourmet guide to edible
org/10.1016/j.foodres.2024.114574. insects: Inner Traditions/Bear.
Rossi, F. (2001). Assessing sensory panelist performance using repeatability and
References reproducibility measures. Food Quality and Preference, 12(5), 467–479. https://doi.
org/10.1016/S0950-3293(01)00038-6
Seo, H., Kim, H. R., & Cho, I. H. (2020). Aroma characteristics of raw and cooked
Alagappan, S., Chaliha, M., Sultanbawa, Y., Fuller, S., Hoffman, L. C., Netzel, G., …
Tenebrio molitor larvae (mealworms). Food Science of Animal Resources, 40(4), 649.
Smyth, H. E. (2021). Nutritional analysis, volatile composition, antimicrobial and
https://doi.org/10.5851/kosfa.2020.e35
antioxidant properties of Australian green ants (Oecophylla smaragdina). Future
Sick, J., Hartmann, A. L., & Frøst, M. B. (2024). Hedonic rating coupled with sensory
Foods, 3, Article 100007. https://doi.org/10.1016/j.fufo.2020.100007
profiles using CATA for six whole roasted or deep-fried insects among Danish 11–13-
Albrektsson, O. (2017). It really bugs me.....: En deskriptiv sensorisk analys av sju ätbara
year-old children. Food Quality and Preference, 114, Article 105094. https://doi.org/
insekter. Bachelor’s Thesis. Örebro University; Örebro, Sweden.
10.1016/j.foodqual.2023.105094
Bae, Y., & Choi, J. (2021). Consumer acceptance of edible insect foods: An application of
Smyth, H. E., Sanderson, J. E., & Sultanbawa, Y. (2012). Lexicon for the sensory
the extended theory of planned behavior. Nutrition Research and Practice, 15(1),
description of Australian native plant foods and ingredients. Journal of Sensory
122–135. Doi: 10.4162%2Fnrp.2021.15.1.122.
Studies, 27(6), 471–481. https://doi.org/10.1111/joss.12012
Chang, C., Wu, G., Zhang, H., Jin, Q., & Wang, X. (2020). Deep-fried flavor:
Suwonsichon, S. (2019). The importance of sensory lexicons for research and
Characteristics, formation mechanisms, and influencing factors. Critical Reviews in
development of food products. Foods, 8(1), 27. https://doi.org/10.3390/
Food Science and Nutrition, 60(9), 1496–1514. https://doi.org/10.1080/
foods8010027
10408398.2019.1575792
Tan, H. S. G., Fischer, A. R. H., Tinchan, P., Stieger, M., Steenbekkers, L. P. A., & van
Dagevos, H. (2021). A literature review of consumer research on edible insects: Recent
Trijp, H. C. M. (2015). Insects as food: Exploring cultural exposure and individual
evidence and new vistas from 2019 studies. Journal of Insects as Food and Feed, 7(3),
experience as determinants of acceptance. Food Quality and Preference, 42, 78–89.
249–259. https://doi.org/10.3920/JIFF2020.0052
https://doi.org/10.1016/j.foodqual.2015.01.013
Deroy, O., Reade, B., & Spence, C. (2015). The insectivore’s dilemma, and how to take
Tuorila, H., & Hartmann, C. (2020). Consumer responses to novel and unfamiliar foods.
the West out of it. Food Quality and Preference, 44, 44–55. https://doi.org/10.1016/j.
Current Opinion in Food Science, 33, 1–8. https://doi.org/10.1016/j.cofs.2019.09.004
foodqual.2015.02.007
Tzompa-Sosa, D. A., Yi, L., Van Valenberg, H., & Lakemond, C. (2019). Four insect oils as
Drake, M. A., & Civille, G. V. (2003). Flavor Lexicons. Comprehensive Reviews in Food
food ingredient: Physical and chemical characterisation of insect oils obtained by an
Science and Food Safety, 2(1), 33–40. https://doi.org/10.1111/j.1541-4337.2003.
aqueous oil extraction. Journal of Insects as Food and Feed, 5(4), 279–292. https://
tb00013.x
doi.org/10.3920/JIFF2018.0020
Elhassan, M., Wendin, K., Olsson, V., & Langton, M. (2019). Quality aspects of insects as
van Boekel, M. A. J. S. (2006). Formation of flavour compounds in the Maillard reaction.
food - nutritional, sensory, and related concepts. Foods, 8(3), 95. https://doi.org/
Biotechnology Advances, 24(2), 230–233. https://doi.org/10.1016/j.
10.3390/foods8030095
biotechadv.2005.11.004
ISO. (2012). Sensory analysis - General guidelines for the selection, training and
van Huis, A. (2013). Potential of insects as food and feed in assuring food security.
monitoring of selected assessors and expert sensory assessors. In (Vol. ISO 8586:
Annual Review of Entomology, 58, 563–583. https://doi.org/10.1146/annurev-ento-
2012).
120811-153704
Jensen, N. H., & Lieberoth, A. (2019). We will eat disgusting foods together - Evidence of
van Huis, A., van Itterbeeck, J., Klunder, H., Mertens, E., Halloran, A., Muir, G., &
the normative basis of Western entomophagy-disgust from an insect tasting. Food
Vantomme, P. (2013). Edible insects: Future prospects for food and feed security. Food
Quality and Preference, 72, 109–115. https://doi.org/10.1016/j.
and Agriculture Organization of the United Nations.
foodqual.2018.08.012
Wassmann, B., Siegrist, M., & Hartmann, C. (2021). Correlates of the willingness to
Jeong, S., & Lee, J. (2021). Effects of cultural background on consumer perception and
consume insects: A meta-analysis. Journal of Insects as Food and Feed, 7, 1–14.
acceptability of foods and drinks: A review of latest cross-cultural studies. Current
https://doi.org/10.3920/JIFF2020.0130
Opinion in Food Science, 42, 248–256. https://doi.org/10.1016/j.cofs.2021.07.004
Wendin, K., Mårtensson, L., Djerf, H., & Langton, M. (2020). Product quality during the
Khatun, H., Claes, J., Smets, R., De Winne, A., Akhtaruzzaman, M., & Van Der Borght, M.
storage of foods with insects as an ingredient: Impact of particle size, antioxidant, oil
(2021). Characterization of freeze-dried, oven-dried and blanched house crickets
content and salt content. Foods, 9(6), 791. https://doi.org/10.3390/foods9060791
(Acheta domesticus) and Jamaican field crickets (Gryllus assimilis) by means of their
Yang, J., & Lee, J. (2019). Application of sensory descriptive analysis and consumer
physicochemical properties and volatile compounds. European Food Research and
studies to investigate traditional and authentic foods: A review. Foods, 8(2), 54. Doi:
Technology, 247(5), 1291–1305. https://doi.org/10.1007/s00217-021-03709-x
10.3390%2Ffoods8020054.
Kiatbenjakul, P., Intarapichet, K. O., & Cadwallader, K. R. (2015). Characterization of
Żołnierczyk, A. K., & Szumny, A. (2021). Sensory and chemical characteristic of two
potent odorants in male giant water bug (Lethocerus indicus Lep. and Serv.), an
insect species: Tenebrio molitor and Zophobas morio larvae affected by roasting
important edible insect of Southeast Asia. Food Chemistry, 168, 639–647. https://doi.
processes. Molecules, 26(9), 2697. https://doi.org/10.3390/molecules26092697
org/10.1016/j.foodchem.2014.07.108
La Barbera, F., Verneau, F., Amato, M., & Grunert, K. (2018). Understanding Westerners’
disgust for the eating of insects: The role of food neophobia and implicit associations.

11