Linköping University | Department of Physics, Chemistry and Biology Type of thesis, 60 hp | Educational Program: Physics, Chemistry and Biology Spring term 2020 | LITH-IFM-x-EX— 20/3809--SE
Taste responsiveness of
black-handed Spider Monkeys (Ateles
geoffroyi) to ten substances
tasting sweet to humans
Sofia Pereira
Examiner, Lina Roth Supervisor, Matthias Laska
Avdelning, institution
Division, Department
Department of Physics, Chemistry and Biology Linköping University
URL för elektronisk version
ISBN
ISRN: LITH-IFM-x-EX--20/3809--SE
_________________________________________________________________
Serietitel och serienummer ISSN
Title of series, numbering ______________________________
Språk Language Svenska/Swedish Engelska/English ________________ Rapporttyp Report category Licentiatavhandling Examensarbete C-uppsats D-uppsats Övrig rapport _____________ Titel Title
Taste responsiveness of black-handed Spider Monkeys (Ateles geoffroyi) to ten substances tasting sweet to humans Författare Author Sofia Pereira Nyckelord Keyword
Ateles geoffroyi, spider monkeys, sweet-tasting substances, taste preference thresholds
Sammanfattning
Abstract
Studies on taste perception in nonhuman primates contribute to the understanding of the evolution of the sense of taste. To assess the responsiveness of four adult spider monkeys (Ateles geoffroyi) to a set of substances perceived as sweet by humans, two-bottle preference tests were performed to determine taste preference thresholds, and taste-induced facial responses were analyzed. The spider monkeys displayed a significant preference for concentrations as low as 0.2-1 mM acesulfame K, 0.002-0.5 mM alitame, 10-20 mM isomalt, 0.002-0.5 mM sodium saccharin, 2-10-20 mM galactose and 10-20-50 mM sorbitol over water. The spider monkeys were generally unable to perceive aspartame and, based on their facial responses, probably do not perceive it as sweet. Thaumatin and monellin were not detected, and most likely neither was the sweetness of sodium cyclamate. Sodium saccharine and sodium cyclamate were rejected at high concentrations by at least one monkey, which is congruent with the perception of a bitter side taste as reported in humans. A significant correlation was found between the ranking order of sweetening potency for the different substances of spider monkeys and humans, but not between spider monkeys and chimpanzees. The results suggest that spider monkeys may be generally more sensitive than chimpanzees and at least as sensitive as humans to the tested substances, supporting the notion that high sensitivity to sweet taste may be associated with a frugivorous dietary specialization. The lack of responsiveness to some of the substances supports the notion of a dichotomy in sweet-taste perception between platyrrhine and catarrhine primates.
Datum
Contents
1. Abstract ... 6
2. Introduction ... 6
3. Materials and Methods ... 9
3.1. Determination of taste preference thresholds in spider monkeys 3.1.1. Animals and housing ... 9
3.1.2. Taste stimuli ... 9
3.1.3. Experimental procedure ... 13
3.1.4. Data analysis ... 14
3.2. Determination of taste detection thresholds in humans 3.2.1. Subjects... 15
3.2.2. Taste stimuli ... 15
3.2.3. Experimental procedure ... 15
3.2.4. Data analysis ... 15
3.3. Analysis of taste-induced facial responses in spider monkeys 3.3.1. Animals and housing ... 16
3.3.2. Taste stimuli ... 16
3.3.3. Experimental procedure ... 17
3.3.4. Data analysis ... 19
4. Results ... 19
4.1. Taste preference thresholds of spider monkeys 4.1.1. Artificial sweeteners ... 19
4.1.2. Sweet-tasting proteins ... 21
4.1.3. Sweet-tasting saccharides ... 22
4.1.4. Interindividual variability ... 23
4.2. Taste detection thresholds of human subjects 4.2.1. Thresholds for alitame and isomalt ... 24
4.2.2. Interindividual variability ... 25
4.3. Taste-induced facial responses in spider monkeys 4.3.1. Inter-rater agreement ... 25
4.3.2. Group-level analysis ... 25
5. Discussion... 31
5.2. Comparison of sweetening potency order between spider monkeys, humans and
chimpanzees... 34
5.3. Between-species comparisons of taste preference thresholds ... 35
5.3.1. Artificial sweeteners ... 35
5.3.2. Sweet-tasting proteins... 42
5.3.3. Sweet-tasting saccharides... 45
5.4. Spider monkeys' perception of aspartame as indicated through facial reactivity ... 47
5.5. Conclusions ... 51
6. Societal and ethical considerations ... 52
7. Acknowledgements ... 53
8. References ... 54
6
1. Abstract
Studies on taste perception in nonhuman primates contribute to the understanding of the evolution of the sense of taste. To assess the responsiveness of four adult spider monkeys
(Ateles geoffroyi) to a set of substances perceived as sweet by humans, two-bottle preference
tests were performed to determine taste preference thresholds, and taste-induced facial responses were analyzed. The spider monkeys displayed a significant preference for concentrations as low as 0.2-1 mM acesulfame K, 0.002-0.5 mM alitame, 10-20 mM isomalt, 0.002-0.5 mM sodium saccharin, 2-20 mM galactose and 20-50 mM sorbitol over water. The spider monkeys were generally unable to perceive aspartame and, based on their facial responses, probably do not perceive it as sweet. Thaumatin and monellin were not detected, and most likely neither was the sweetness of sodium cyclamate. Sodium saccharine and sodium cyclamate were rejected at high concentrations by at least one monkey, which is congruent with the perception of a bitter side taste as reported in humans. A significant correlation was found between the ranking order of sweetening potency for the different substances of spider monkeys and humans, but not between spider monkeys and chimpanzees. The results suggest that spider monkeys may be generally more sensitive than chimpanzees and at least as sensitive as humans to the tested substances, supporting the notion that high sensitivity to sweet taste may be associated with a frugivorous dietary specialization. The lack of responsiveness to some of the substances supports the notion of a dichotomy in sweet-taste perception between platyrrhine and catarrhine primates.
Keywords: Ateles geoffroyi, spider monkeys, sweet-tasting substances, taste preference thresholds
2. Introduction
Comparative studies of taste perception are an important tool to better understand the mechanisms underlying the evolution of the sense of taste. Taste allows responses to different taste qualities which supply valuable sensory input for an animal’s decision of what to ingest (Nelson et al. 2001). For example, bitter receptors trigger behavioral responses of aversion towards potentially toxic compounds whereas sweet receptors allow for the recognition of highly caloric food resources (Nelson et al. 2001). However, different species have been found
7
to vary considerably in their taste perception for the same substances, which is thought to reflect evolutionary adaptations to a specific diet (Breslin 2013).
Primates are a particularly suitable order of mammals for the study of taste perception as they comprise a large variety of dietary specializations and the composition of food is commonly thought to affect the taste perception of a given species (Dominy et al. 2001). Black-handed spider monkeys (Ateles geoffroyi) are highly frugivorous New World primates specialized on consuming ripe fruit and their food selection behavior suggests that they may use the sweetness of fruits as a criterion for consumption (Di Fiore et al 2008; Gonzalez-Zamora et al. 2009). Their responsiveness to sweet-tasting carbohydrates (Laska et al. 1996, 1998, 2001) as well as to sour (Laska et al. 2000, 2003), bitter (Laska et al. 2009), salty, and umami (Laska and Hernandez Salazar 2004; Laska et al. 2008) taste stimuli has been assessed in previous studies.
In addition to the carbohydrates sucrose, fructose, glucose, maltose, and lactose which are commonly found in the food of primates (Nagy and Shaw 1980; Kinghorn and Soejarto 1986), a wide variety of substances from diverse chemical classes are perceived as “sweet” by humans (Marie and Piggott 1991; Merillon and Ramawat 2018). Structurally, these sweet-tasting substances range from simple amino acids over peptides and proteins to terpenoids, flavonoids, steroidal saponins, and dihydrochalcones, to name but a few chemical classes. Considering that mammals have only one type of sweet-taste receptor, the TAS1R2+TAS1R3 heterodimer receptor (Bachmanov and Beauchamp 2007), it is intriguing that one type of receptor is capable of binding such a wide array of structurally diverse ligands, though at different affinities. Recent studies have found that both allelic and copy number variation in the sweet-taste receptor gene may explain the marked differences within and between species that have been reported in the ability to perceive and in the sensitivity for sweeteners (Dias et al. 2015). In addition to their chemical nature, sweeteners can be subdivided into naturally occurring ones which are usually plant-derived, and artificial ones which are not found in nature. This distinction may be interesting to study with regard to taste perception as the former group of substances may share a long evolutionary history with a given species whereas the latter does not. This, in turn, may have implications for the detectability and acceptance of sweet-tasting substances by an animal. The artificial sweeteners aspartame and neotame, for example, have been reported to be detectable by catarrhine primates, but not by prosimian or platyrrhine primates (Glaser et al. 1992).
8
Furthermore, sweeteners can be classified according to their sweetening potency. So-called
high-potency sweeteners are usually perceived by humans as considerably more intensely
sweet compared to an isomolar sucrose solution whereas low-potency sweeteners are accordingly perceived as less intensely sweet than sucrose. Very little so far is known about whether nonhuman primates perceive sweeteners in the same manner as humans with regard to their sweetening potency.
Finally, sweet-tasting substances can also be classified based on their caloric values. Ingestion of non-caloric sweeteners provide, as the name implies, no calories at all whereas so-called
sugar substitutes provide a lower amount of calories relative to common carbohydrates.
Taste preference thresholds, as a first and widely used approximation of an animal’s taste sensitivity, have been determined in a variety of primate species from all major taxa (e.g. strepsirrhines: Wielbass et al. 2015; platyrrhines: Laska et al. 1996; catarrhines: Laska et al. 1999; Laska 2000). Such thresholds correspond to the lowest concentration of a given substance an animal prefers over water. However, most studies so far only used prototypical representatives of the five basic taste qualities, e.g. quinine as the prototypical representative of bitter taste. With regard to sweet-tasting substances, most studies only assessed the sensitivity of primates for the five food-associated carbohydrates sucrose, fructose, glucose, maltose, and lactose. Surprisingly little, in contrast, is known so far about the sensitivity of the sense of taste in primates at the behavioral level for sweet-tasting substances other than carbohydrates.
Thus, in order to gain more insight into the sweet-taste perception of nonhuman primates, and into the notion that dietary specialization rather than phylogenetic relatedness may account for between-species differences in sweet-taste perception, the present study aimed at assessing the taste responsiveness of spider monkeys to a diverse set of substances perceived as sweet by humans. These include substances from widely different chemical classes, as well as both
high-potency and low-high-potency sweeteners (as perceived by humans), artificial and naturally occurring sweeteners, as well as caloric and non-caloric sweeteners. Thresholds of different
primate and non-primate species were compared for phylogenetic relevance. Furthermore, in an attempt to better understand whether the spider monkeys, similar to human subjects, may perceive an unpleasant side taste with some of the artificial sweeteners, taste-induced facial responses were video-recorded and analyzed.
9
3. Materials and Methods
3.1. Determination of taste preference thresholds in spider monkeys
3.1.1. Animals and housing
The present study included four adult black-handed spider monkeys (Ateles geoffroyi). The monkeys were housed at the Universidad Veracruzana’s research station UMA Doña Hilda Ávila de O’Farrill near Catemaco, Mexico. The group of subjects was composed of one female, Mari, and three males, Cejitas, Gruñon and Lucas. The female was fifteen years old, and the males were fifteen, twelve and fifteen years old, respectively. The four individuals were not genetically related with each other. All individuals were housed in a series of enclosures exposed to natural light and connected to each other through sliding doors. The monkeys were fed a wide variety of fresh seasonal fruits and vegetables once a day.
3.1.2. Taste stimuli
A set of ten substances tasting sweet to humans was used (Figure 1). A sweetener can be described as a ‘high-potency’ sweetener or as a ‘low-potency’ sweetener according to its sweetening potency relative to sucrose. Here, these designations are employed in accordance with the definition by which a high-potency sweetener has a sweetening potency of at least a factor of 10 higher than the one of sucrose and a low-potency sweetener has a sweetening potency lower than sucrose (Yasuura 2014).
Acesulfame K (CAS# 55589-62-3)
An artificial, high-potency sweetener perceived by humans as about 200 times sweeter than sucrose. It is a non-caloric sweetener containing sulphur and nitrogen, and chemically the potassium salt of an oxathiazine. Reported to have a bitter-side taste at high concentrations.
Alitame (CAS# 99016-42-9)
An artificial, high-potency and non-caloric sweetener perceived by humans as about 2,000 times sweeter than sucrose. Chemically a dipeptide composed of L-aspartic acid and L-alanine.
10
Aspartame (CAS# 22839-47-0)
An artificial, high-potency and non-caloric sweetener perceived by humans as about 200 times sweeter than sucrose. Chemically a methyl ester of the dipeptide built by the amino acids L-aspartic acid and L-phenylalanine. Reported to have a bitter side taste at high concentrations.
Isomalt (CAS# 64519-82-0)
An artificial low-potency sweetener perceived by humans as about 45% as sweet as sucrose. It is a nutritive sweetener, providing 2.0 kcal/g, half the energy value provided by sucrose. Chemically a sugar alcohol manufactured from sucrose.
Sodium cyclamate (CAS# 139-05-9)
An artificial high-potency sweetener perceived by humans as about 30-50 times sweeter than sucrose. Chemically the sodium salt of cyclamic acid. Reported to have a sour side taste at high concentrations.
Sodium saccharine (CAS# 81-07-2)
An artificial, high-potency and non-caloric sweetener perceived by humans as about 300-400 times sweeter than sucrose. Chemically the sodium salt of benzoic sulfimide and thus a sulphur- and nitrogen-containing compound. Reported to have a bitter side taste at high concentrations.
Monellin (CAS# 9062-83-3)
A naturally occurring high-potency sweetener perceived by humans as about 800-2,000 times sweeter than sucrose. Chemically a protein found in the West African serendipity berry
(Dioscoreophyllum cumminsii).
Thaumatin (CAS# 53850-34-3)
A naturally occurring high-potency sweetener perceived by humans as about 2,000 times sweeter than sucrose. Chemically a protein found in the West African katemfe fruit
(Thaumatococcus daniellii).
Galactose (CAS# 59-23-4)
A naturally occurring low-potency sweetener perceived by humans as about 30% as sweet as sucrose. Chemically a monosaccharide and a constituent of the disaccharide lactose.
11
Sorbitol (CAS# 50-70-4)
A naturally occurring low-potency sweetener perceived by humans as about 60% as sweet as sucrose. A nutritive sweetener, providing 2.6 kcal/g. Chemically a sugar alcohol found in certain fruits.
Acesulfame K and aspartame were obtained from Ter Hell & Co., Hamburg, Germany, thaumatin was obtained from Xi’an Sgonek Biological Technology Co. Ltd., Xi’an, China and isomalt was obtained from Beneo-Palatinit, Mannheim, Germany. All other substances were obtained from Sigma-Aldrich, Stockholm, Sweden. All tested substances were of the highest available purity (>99%).
12
Acesulfame K Alitame Aspartame
Figure 1: The sweet-tasting substances tested in the present study. For all substances but Monellin and Thaumatin, a 2-D representation of their molecular structure is presented. Monellin and Thaumatin are described in 3-D ribbon diagrams.
(National Centre for Biotechnology Information)
Sorbitol Galactose Thaumatin Monellin Sodium saccharin Sodium cyclamate Isomalt
13
3.1.3. Experimental procedure
Two-bottle preference tests of short duration (Richer & Campbell 1940) were performed to determine the spider monkeys’ taste preference thresholds to the different sweeteners. The animals were simultaneously presented with two 100 mL graduated plastic cylinders with metal drinking spouts, one containing tap water and the other containing an aqueous solution of the taste stimulus at a defined concentration (Figure 2). The individuals were allowed 1 minute to drink from the bottles. In each trial, the experimenter made sure to observe that the animals had tried both bottles. After each trial, the amount of liquid consumed from each bottle was recorded. The testing occurred in the morning before the monkeys were fed, and up to six trials per monkey were performed per day, with inter-trial intervals of at least 10 minutes in order to ensure that no interference from the previous trial occurred. The data collection began in May 2019 and ended in October 2019.
Figure 2: The two-bottle preference test
With each substance, testing started with a presumably perceptible and thus attractive concentration of a given taste stimulus. The starting concentrations were determined as approximately a factor of 10-100 above the known human taste detection threshold value (van
14
Gemert 2011). The initial concentration was then decreased in 10-fold steps until the animals failed to show a preference. In case the subjects did not show a preference for the initial concentration, gradually higher concentrations were tested in a 10-fold step fashion until a preference was finally shown or until a concentration range covering three 10-fold dilution steps was tested. In order to determine the monkeys’ taste preference thresholds with higher precision, two intermediate concentrations which were between the highest concentration for which an animal did not show a preference and the lowest concentration for which the animal showed a preference were tested. The taste substances were tested sequentially, meaning that testing with a novel substance only started when an animal had finished the testing with a previous substance.
Each concentration of a given taste substance was tested ten times per animal. In order to reduce any potential bias in testing as well as to keep the animals motivated, different concentrations of a given taste substance were presented in a pseudo-randomized order. The same concentration was never consecutively presented to the animals for more than three trials in a row. Additionally, in order to control for any potential side-bias, the side at which the bottle containing the sweet-tasting substance was presented was also pseudo-randomized. This ensured that the sweet-tasting solution was positioned on the left the same number of times as it was positioned on the right, without ever being presented on the same side more than three consecutive times.
3.1.4. Data analysis
The quantities of sweet-tasting solution and tap water consumed over the ten trials were summed and converted to percentages relative to the total amount of liquid consumed. An individual was considered to have preferred the sweet-tasting substance over water if the percentage of consumed sweet-tasting solution was at least 66.7% of the total liquid consumed. If an individual consumed less than 33.3% of the sweet-tasting solution relative to the total amount of liquid, it was considered to have rejected the sweet-tasting substance. Additionally, a one-tailed binomial test was performed. Thus, in addition to the criterion of having reached 66.7% of consumed sweet-tasting substance, an individual should also have drunk more of the sweet-tasting solution than water in at least 8 out of 10 trials (p < 0.05) with a given concentration of a given taste substance to be considered as significantly preferred over water. The taste preference threshold for each substance was considered as the lowest concentration significantly preferred over water by each individual.
15
The Spearman rank-order correlation test was performed to evaluate the correlation between the ranking order of sweetening potencies of spider monkeys, chimpanzees and humans.
3.2. Determination of taste detection thresholds in humans
As the sensitivity of humans towards alitame and isomalt had never been tested, I decided to determine the human taste detection thresholds for these two substances in order to obtain a means of comparison for the taste preference thresholds determined for the spider monkeys.
3.2.1. Subjects
Ten human subjects (Homo sapiens) between the ages of 23 and 26 years old were tested. Of the ten individuals included, five were women and five were men.
3.2.2. Taste stimuli
The tested taste stimuli were alitame and isomalt. For a detailed description of these substances please consult the one provided in the previous section (3.1.2. Taste stimuli).
3.2.3. Experimental procedure
An up-down, two-alternative forced choice (2-AFC) staircase procedure was performed to determine human taste detection thresholds (Snyder et al. 2015). According to this method, each subject was asked to indicate which of two presented solutions contained the taste stimulus. The testing started at a clearly detectable concentration (0.05 mM for alitame and 160 mM for isomalt), which was then decreased in 2-fold concentration steps until the subject failed to detect the substance. Each concentration was presented twice to the subject. Two correct choices resulted in a subsequent decrease in concentration in the following set of two trials. An incorrect choice, i.e. failing to correctly identify the solution containing the taste stimulus, was followed by an increase in concentration. Once the testing shifted from a decreasing concentration fashion to an increasing concentration fashion, a reversal was registered. The testing terminated once a subject reached seven reversals, i.e. seven turning points in the direction of the concentration staircase.
All subjects were asked not to eat, chew or smoke in the ten minutes preceding the testing session to ensure that no previous taste lingering in the mouth interfered with the subject’s performance. Tap water was provided for the subjects to rinse their mouth in between each moment of tasting the test solutions.
16
3.2.4. Data analysis
According to the previously described method, each reversal registered corresponds to a tested concentration. Thus, each reversal step was registered and ordered ascendingly, in order to find the median reversal value. The taste detection threshold of a subject corresponds to the concentration value above the one identified as the median value.
3.3. Analysis of taste-induced facial responses in spider monkeys
3.3.1. Animals and housing
The present experiment included six adult black-handed spider monkeys (Ateles geoffroyi) housed at the Universidad Veracruzana’s research station UMA Doña Hilda Ávila de O’Farrill near Catemaco, Mexico. The monkeys were kept in the same conditions previously described in this report. The group of subjects was composed of three females, Frida, Margarita and Mari, and three males, Cejitas, Gruñon and Lucas. The females were twelve, thirteen and fifteen years old, and the males were fifteen, twelve and fifteen years old, respectively. The group of animals included those who had participated in the two-bottle preference tests.
3.3.2. Taste stimuli
A total of five different taste substances was presented to the spider monkeys. Three of the substances represent different taste qualities. Sucrose was used as a sweet taste stimulus,
caffeine was used as a bitter taste stimulus and citric acid was used as a sour taste stimulus. Tap water was used as a neutral stimulus and thus, a control. The fifth stimulus, aspartame,
was included as a means to better comprehend the monkey’s perception of the substance and thus the results of the previously conducted two-bottle preference tests. In the previously described two-bottle preference tests, two out of the four individuals presented with aspartame showed a preference for highest tested concentration (20 mM). In contrast, the remaining two individuals rejected the same concentration, which is consistent with the notion of a bitter side-taste of aspartame as perceived by humans (Schiffman et al. 1995). Sucrose, caffeine and citric acid were presented at a concentration of 200 mM, 100 mM, and 500 mM, respectively, in order to provide taste stimuli that are clearly detectable for the animals, but not overwhelmingly sweet, bitter, and sour, respectively. Aspartame was used at the concentration of 100 mM, for being well above the previously tested concentration (20 mM), which two of the spider monkeys preferred over water. Table 1 describes the concentrations used for each of the taste stimuli.
17
Table 1. Concentrations in mM of the different taste stimuli presented to the spider monkeys.
3.3.3. Experimental procedure
Each substance was presented to the animals using a glass dropper contained in a dark-glass container (Figure 3). Prior to any critical tests, the animals were gradually made familiar with this device through several trials using sucrose as an attractive and thus motivational stimulus. Once the animals were completely familiarized with the device, each substance was presented ten times to each individual. A trial was considered each time a substance was presented to an individual who clearly tasted and/or ingested the taste stimulus. The tasting or ingestion of the taste stimulus was recorded with a video camera (Sony Handycam HDR-CX405) until the dropper (2.5 ml) was emptied or until the animal refused to take in any more substance. In order to prevent the animals from associating any visual cues to the different taste substances, all droppers and containers were of the same model and displayed no visible differences between them. All taste solutions were colorless, ensuring that the animals were not able to pick up on any discriminatory indications between substances. Three to six trials were performed per day with the same substance, depending on the experimenter’s schedule and the animals’ motivation to cooperate The taste substances were tested sequentially, meaning that testing with a novel substance only started when an animal had finished the testing with a previous substance. However, trials with presumably aversive taste stimulus (such as caffeine and citric acid) were sometimes interspersed with sucrose trials to maintain the animals motivated to cooperate. In any case, inter-trial intervals of at least 10 minutes were implemented. Stimulus Concentration (mM) Sucrose 200 Caffeine 100 Citric Acid 500 Aspartame 100
18
Figure 3: Transparent-glass dropper and respective dark-glass container used to present the different taste stimuli to the spider monkeys.
Once all trials were completed, two coders experienced in working with spider monkeys were asked to analyze the frequency of a list of previously selected facial expressions and facial motor patterns considered relevant for the study. Both coders were unaware of which substance was being presented to the animal in each video. The ethogram compiling the list of behaviors recorded in each trial and its respective description is displayed in the Appendix section of this report. The selected facial expressions had been previously found to be indicative of the pleasantness or unpleasantness of taste stimuli as perceived by several non-human primate species including great apes, Old World monkeys, New World monkeys and prosimians (Steiner et al. 2001). Even though only behavioral frequency was analyzed, the behaviors were separated into two categories. Behaviors which could be analyzed by both frequency and duration were categorized as state behaviors. Behaviors which could only be analyzed by frequency were categorized as behavioral events.
19
3.3.4. Data analysis
As a preliminary validation of the data and in order to verify a sufficient level of agreement between the two coders’ independent analysis of the video footage, an Intraclass Correlation Analysis (ICC) was run. In order for the data to be considered reliable and included in the study, a correlation coefficient (α) of at least 0.7 was necessary.
The data points which achieved sufficient inter-rater agreement were further analyzed for significant differences in facial responses between the different tested substances. This was done by performing the Related-Samples Wilcoxon Signed Rank test for all pairwise comparisons between the different stimuli for each of the selected behaviors at the group level, pulling all individuals together.
4. Results
4.1. Taste preference thresholds of spider monkeys
4.1.1. Artificial sweeteners
For acesulfame K the four spider monkeys displayed taste preference thresholds between 0.2 mM and 1 mM. The taste preference threshold was 0.2 mM for Lucas, 0.5 mM for Cejitas and Mari, and 1mM for Gruñon (p<0.05, binomial test) (Figure 4). The taste preference thresholds for alitame ranged from 0.002 mM to 0.005 mM. Lucas and Cejitas displayed a taste preference threshold of 0.002 mM and Mari and Gruñon displayed a taste preference threshold of 0.005 mM (p<0.05, binomial test) (Figure 4). The taste preference threshold for aspartame was 20 mM for Lucas and Mari (p<0.05, binomial test) (Figure 4). Cejitas and Gruñon did not show a preference for this substance at any of the concentrations tested. Rather, Cejitas rejected the concentrations 20, 2 and 0,02 mM, and Gruñon rejected the concentrations 20, 2 and 0,2 mM. For isomalt, the taste preference threshold was 10 mM for Mari and 20 mM for Cejitas, Gruñon and Lucas (p<0.05, binomial test) (Figure 4). The taste preference threshold for sodium
cyclamate was 1 mM for Cejitas (p<0.05, binomial test) (Figure 4). None of the three
remaining individuals displayed a preference for this substance at any of the concentrations tested. Rather, Mari and Lucas rejected the substance at the concentrations 10 and 1 mM, and Gruñon rejected the substance at the concentration 10 mM. For sodium saccharine, the taste preference thresholds ranged from 0.002 mM to 0.5 mM. The taste preference thresholds for
20
this substance were 0.002 mM for Lucas, 0.02 mM for Cejitas, 0.2 mM for Mari and 0.5 mM for Gruñon (p<0.05, binomial test) (Figure 4).
0 10 20 30 40 50 60 70 80 90 100 0.1 1 10 P refer en ce fo r aces u lf ame K (%) Concentration (mM)
Acesulfame K
0 10 20 30 40 50 60 70 80 90 100 0.001 0.01 0.1 1 10 P refer en ce fo r al it ame (%) Concentration (mM)Alitame
0 10 20 30 40 50 60 70 80 90 100 0.02 0.2 2 20 P refer en ce fo r as p ar tame (%) Concentration (mM)Aspartame
0 10 20 30 40 50 60 70 80 90 100 5 50 500 P refer en ce fo r is o mal t (%) Concentration (mM)Isomalt
21
Figure 4: The taste responsiveness of the four spider monkeys when presented simultaneously with water and the artificial sweeteners acesulfame K, alitame, aspartame, isomalt, sodium cyclamate and sodium saccharin. The blue triangles, orange squares, yellow rhombi and green circles represent data points for Mari, Lucas, Cejitas and Gruñon, respectively. The horizontal dashed lines at 33.3%, 50% and 66.7% represent the criterion of the rejection, the level of chance and the criterion of preference, respectively.
4.1.2. Sweet-tasting proteins
For monellin, the taste preference threshold was 0.001 mM for Mari (p<0.05, binomial test) (Figure 5). None of the other individuals showed a preference for this substance. Cejitas and Gruñon displayed a rejection for the substance at the concentrations 0.01, 0.001 and 0.0001 mM, and Lucas displayed a rejection at 0.01 and 0.001 mM. For thaumatin, only Mari displayed a preference, with a taste preference threshold of 0.1 mM (p<0.05, binomial test) (Figure 5). Cejitas rejected the concentrations 0.1 and 0.01 mM, and Gruñon rejected 0.1, 0.01, 0.001 and 0.0001 mM. Lucas did not display a preference for thaumatin nor rejected the substance. 0 10 20 30 40 50 60 70 80 90 100 0.1 1 10 P refer en ce fo r so d iu m cy cl ama te (%) Concentration (mM)
Sodium cyclamate
0 10 20 30 40 50 60 70 80 90 100 0.001 0.01 0.1 1 10 P refer en ce fo r sacc h ar in (%) Concentration (mM)Saccharin
22
Figure 5: The taste preference thresholds of the four spider monkeys when presented simultaneously with water and the sweet-tasting proteins monellin and thaumatin. The blue triangles, orange squares, yellow rhombi and green circles represent data points for Mari, Lucas, Cejitas and Gruñon, respectively. The horizontal dashed lines at 33.3%, 50% and 66.7% represent the criterion of the rejection, the level of chance and the criterion of preference, respectively.
4.1.3. Sweet-tasting saccharides
The taste preference thresholds for galactose ranged between 2 mM and 20 mM in the group of spider monkeys. The taste preference threshold was 2 mM for Cejitas and 20 mM for Gruñon, Lucas and Mari (p<0.05, binomial test) (Figure 6). For sorbitol, the taste preference thresholds were 20 mM for Cejitas, Lucas and Mari, and 50 mM for Gruñon (p<0.05, binomial test) (Figure 6). 0 10 20 30 40 50 60 70 80 90 100 0.00001 0.0001 0.001 0.01 P refer en ce fo r mo n el lin (%) Concentration (mM)
Monellin
0 10 20 30 40 50 60 70 80 90 100 0.0001 0.001 0.01 0.1 P refer en ce fo r th au mat in (%) Concentration (mM)Thaumatin
23
Figure 6: The taste preference thresholds of the four spider monkeys when presented simultaneously with water and the saccharides galactose and sorbitol. The blue triangles, orange squares, yellow rhombi and green circles represent data points for Mari, Lucas, Cejitas and Gruñon, respectively. The horizontal dashed lines at 33.3%, 50% and 66.7% represent the criterion of the rejection, the level of chance and the criterion of preference, respectively.
4.1.4. Interindividual variability
With four of the ten taste substances, inter-individual variability in taste thresholds was low. The difference between the most- and least sensitive animal ranged from a dilution factor of only 2 with isomalt, a dilution factor of 2.5 with sorbitol and alitame, to a dilution factor of 5 with acesulfame K. For galactose, the difference between the lowest and highest taste preference thresholds determined was a dilution factor of 10, with only one monkey differing from the others. The highest interindividual variability was observed with sodium saccharin for which each monkey displayed a different taste preference threshold. For the latter substance the difference between the most- and least-sensitive animal was a dilution factor of 250. For
aspartame, the only two individuals who showed a preference for the substance, Mari and
Lucas, displayed the same taste preference threshold. For monellin and thaumatin, only Mari showed a preference for the substances. For sodium cyclamate, only Cejitas preferred the substance over water. In general, Lucas showed a higher sensitivity than the other monkeys, as he displayed the lowest taste preference thresholds with three out of the six substances that all animals preferred over water.
0 10 20 30 40 50 60 70 80 90 100 1 10 100 P refer en ce fo r gal act o se (%) Concentration (mM)
Galactose
0 10 20 30 40 50 60 70 80 90 100 10 100 Pr ef er en ce fo r so rb it o l ( % ) Concentration (mM)Sorbitol
24
4.2. Taste detection thresholds of human subjects
4.2.1. Thresholds for alitame and isomalt
For alitame, the human subjects displayed thresholds ranging between 0.0015625 mM and 0.0125 mM (Figure 7). Only one of the subjects displayed a taste detection threshold of 0.0015625 mM. Five subjects displayed a taste detection threshold of 0.00625 mM and for the remaining four subjects the threshold values were of 0.0125 mM. Accordingly, the median value for the human taste detection threshold for alitame was 0.00625 mM.
For isomalt, the human subjects displayed thresholds ranging between 5 mM and 40 mM (Figure 7). Only one of the subjects displayed a taste detection threshold of 5 mM. Similarly, only one of the subjects showed a taste detection threshold of 10 mM. Six subjects displayed a taste detection threshold of 20 mM and for the remaining two individuals displayed a threshold of 40 mM. Accordingly, the median value for the human taste detection threshold for alitame was 20 mM.
Figure 7: Taste detection thresholds of the ten human subjects when presented simultaneously with water and the artificial sweeteners alitame and isomalt. Each dot represents a human subject. The horizontal lines represent the median value.
0 5 10 15 20 25 30 35 40 45 Tas te d et ect io n t h res h o ld (mM ) Subjects
Isomalt
0.000 0.002 0.004 0.006 0.008 0.010 0.012 0.014 Tas te d et ect io n t h res h o ld (mM ) SubjectsAlitame
25
4.2.2. Interindividual variability
For both alitame and isomalt, there was considerable variation in sensitivity across individuals. The difference between the most- and least sensitive subject for both substances was a dilution factor of 8.
4.3. Taste-induced facial responses in spider monkeys 4.3.1. Inter-rater agreement
Inter-rater agreement was sufficient (α ≥ 0.7, ICC) for all tested substances with four of the six individuals included in the experiment. These individuals were Frida, Margarita, Gruñon and Lucas, of which only Gruñon and Lucas had also previously participated in the two-bottle preference tests. The data relative to the remaining two individuals, Cejitas and Mari, achieved a sufficient level of inter-rater agreement for only three and four of the tested substances, respectively, and were thus removed from the following analysis.
4.3.2. Group-level analysis
For licking, significant differences were found for all pairwise comparisons between the different stimuli (Aspartame VS Caffeine: Z = -4.147, p < 0.001; Aspartame VS Citric Acid: Z = -4.288, p < 0.001; Aspartame VS Water: Z = 3.137, p < 0.01; Aspartame VS Sucrose: Z = 5.321, p < 0.001; Sucrose VS Water: Z = 4.722, p < 0.001 ; Sucrose VS Citric Acid: Z = 5.523, p < 0.001; Sucrose VS Caffeine: Z = 5.437, p < 0.001; Citric Acid VS Caffeine: Z = -2.161, p < 0.05; Citric Acid VS Water: Z = -5.016, p < 0.001; Caffeine VS Water: Z = -4.798, p < 0.001; Wilcoxon). Licking occurred more often during sucrose, followed by control and aspartame trials, and only then by caffeine and citric acid trials (Figure 8a).
Sucking frequency was significantly higher during sucrose trials than during caffeine trials (Z
= -2.636, p < 0.01, Wilcoxon). No other pairwise comparison between any other of the different tested substances revealed a significant difference (Aspartame VS Caffeine: Z = -1.342, p > 0.05; Aspartame VS Citric Acid: Z = -0.535, p > 0.05; Aspartame VS Water: Z = -0.378, p > 0.05; Aspartame VS Sucrose: Z = 1.345, p > 0.05; Sucrose VS Water: Z = -1.31, p > 0.05; Sucrose VS Citric Acid: Z = -1.653, p > 0.05; Citric Acid VS Caffeine: Z = 1, p > 0.05; Citric Acid VS Water: Z =0.816, p > 0.05; Caffeine VS Water: Z = 1.633, p > 0.05; Wilcoxon). Regarding sniffing frequency, all substances differed significantly from each other (Aspartame VS Water: Z = 3.592, p < 0.001; Aspartame VS Sucrose: Z = -3.491, p < 0.001; Sucrose VS Water: Z = 4.704, p < 0.001 ; Sucrose VS Caffeine: Z = 3.8, p < 0.001; Citric Acid VS Caffeine:
26
Z = -2.078, p < 0.05; Citric Acid VS Water: Z = -3.796, p < 0.001; Caffeine VS Water: Z = 3.292, p < 0.01; Wilcoxon) with the exception of the pairwise comparisons between sucrose and citric acid (Z = 1.811, p > 0.05, Wilcoxon), and aspartame and caffeine (Z = 0.068, p > 0.05, Wilcoxon) (Figure 8b). A trend rather than a significant difference was found for the comparison between aspartame and citric acid (Z = -1.908, p = 0.056, Wilcoxon).
With regard to the state of the eyes, the frequency of eyes open more than 50% differed only in the pairwise comparisons between aspartame and citric acid (Z = -2.214, p < 0.05, Wilcoxon), caffeine and water (Z = 2.674 , p < 0.01, Wilcoxon), and citric acid and water (Z = -3.037, p < 0.01, Wilcoxon). A trend rather than a significant difference was found for the comparison between aspartame and caffeine (Z = -1.768, p = 0.077, Wilcoxon). No significant differences were found for the pairwise comparisons between aspartame and water (Z = 0.939, p > 0.05, Wilcoxon), aspartame and sucrose (Z = -0.228, p > 0.05, Wilcoxon), sucrose and caffeine (Z = -1.12, p > 0.05, Wilcoxon), sucrose and citric acid (Z = -1.238, p > 0.05, Wilcoxon), sucrose and water (Z = 0.986, p > 0.05, Wilcoxon), and caffeine and citric acid (Z = -0.312, p > 0.05, Wilcoxon). Higher frequencies of this behavior were recorded for water, and the lower for caffeine and citric acid.
Frequencies of eyes open less than 50% differed significantly between all substances (Aspartame VS Caffeine: Z = -2.8, p < 0.01; Aspartame VS Citric Acid: Z = -3.438, p < 0.01; Aspartame VS Water: Z = 2.657, p < 0.01; Aspartame VS Sucrose: Z = 3.429, p < 0.01; Sucrose VS Citric Acid: Z = -4.596, p < 0.001; Sucrose VS Caffeine: Z = -4.64, p < 0.001; Citric Acid VS Water: Z = -4.878, p < 0.001; Caffeine VS Water: Z = 4.574, p < 0.001; Wilcoxon) with the exception of the pairwise combinations between caffeine and citric acid (Z = -0.869, p > 0.05, Wilcoxon) and tap water and sucrose (Z = -1.182, p > 0.05, Wilcoxon) (Figure 8c). In general, eyes were open less than 50% the most during sucrose and control ingestions.
The frequency of closed eyes differed significantly between aspartame and caffeine (Z = -2.419, p < 0.05, Wilcoxon), aspartame and citric acid (Z = -2.247, p < 0.05, Wilcoxon), aspartame and sucrose (Z = 2.965, p < 0.01, Wilcoxon), sucrose and caffeine (Z = -3.767, p < 0.001, Wilcoxon), sucrose and citric acid (Z = -3.864, p < 0.001, Wilcoxon), and sucrose and tap water (Z = -3.05, p < 0.01, Wilcoxon) (Figure 8d). No other significant differences were found (Aspartame VS Water: Z = -0.789, p > 0.05; Caffeine VS Water: Z = 1.375, p > 0.05; Citric Acid VS Water: Z = -1.368, p > 0.05; Citric Acid VS Caffeine: Z = -0.237, p > 0.05;
27
Wilcoxon). In general, eyes were closed the most during sucrose ingestions, followed by aspartame ingestions. Eyes were closed the least during citric acid ingestions.
The frequency of flat tongue protrusions differed significantly between all substances (Aspartame VS Caffeine: Z = -3.520, p < 0.001; Aspartame VS Citric Acid: Z = -3.468, p < 0.001; Aspartame VS Water: Z = 3.378, p < 0.01; Aspartame VS Sucrose: Z = 5.238, p < 0.001; Sucrose VS Water: Z = -4.323, p < 0.001 ; Sucrose VS Citric Acid: Z = -5.509, p < 0.001; Sucrose VS Caffeine: Z = -5.407, p < 0.001; Citric Acid VS Water: Z = -4.873, p < 0.001; Caffeine VS Water: Z = 4.641, p < 0.001; Wilcoxon) with the exception of the pairwise comparison between caffeine and citric acid, for which a trend rather than a significant difference was found (Z = -1.853, p = 0.064, Wilcoxon) (Figure 8e). In general, flat tongue protruding occurred the most during sucrose ingestions, followed by control and then aspartame ingestions. Flat tongue protruding was less frequent during caffeine and citric acid ingestions.
No significant differences were found between substances regarding the frequency of
downward-directed tongue protrusions (Aspartame VS Caffeine: Z = -0.977, p > 0.05;
Aspartame VS Citric Acid: Z = -0.447, p > 0.05; Aspartame VS Water: Z = -1.319, p > 0.05; Aspartame VS Sucrose: Z = -0.978, p > 0.05; Sucrose VS Water: Z = -0.542, p > 0.05; Sucrose VS Citric Acid: Z = 0.832, p > 0.05; Sucrose VS Caffeine: Z = 0.306, p > 0.05; Citric Acid VS Caffeine: Z = 0.884, p > 0.05; Citric Acid VS Water: Z = 1.327, p > 0.05; Caffeine VS Water: Z = -0.952, p > 0.05, Wilcoxon) nor upward-directed tongue protrusions (Aspartame VS Caffeine: Z = -1, p > 0.05; Aspartame VS Citric Acid: Z = -1, p > 0.05; Aspartame VS Water: Z = 0.816, p > 0.05; Aspartame VS Sucrose: Z = 1.604, p > 0.05; Sucrose VS Water: Z = 0,816, p > 0.05; Sucrose VS Citric Acid: Z = 1.604, p > 0.05; Sucrose VS Caffeine: Z = -1.604, p > 0.05; Citric Acid VS Caffeine: Z = -1, p > 0.05; Citric Acid VS Water: Z = -1.342, p > 0.05; Caffeine VS Water: Z = -1.342, p > 0.05, Wilcoxon).
The frequency of repetitive tongue protruding differed significantly between caffeine and citric acid (Z = 2.309 , p < 0.05, Wilcoxon), aspartame and tap water (Z = -2, p < 0.05, Wilcoxon), sucrose and tap water (Z = -2, p < 0.05, Wilcoxon), and tap water and citric acid (Z = 2.714, p < 0.01, Wilcoxon). No other significant differences were found (Aspartame VS Caffeine: Z = -1.342, p > 0.05; Aspartame VS Citric Acid: Z = 1.890, p > 0.05; Aspartame VS Sucrose: Z < 0.001, p > 0.05; Sucrose VS Citric Acid: Z = 1.291, p > 0.05; Sucrose VS Caffeine: Z = -1.342, p > 0.05; Caffeine VS Water: Z = -1, p > 0.05, Wilcoxon).
28
The frequency of tongue protrusion gape differed significantly between all tested substances (Aspartame VS Caffeine: Z = -4.063, p < 0.001; Aspartame VS Citric Acid: Z = -4.138, p < 0.001; Aspartame VS Water: Z = 3.226, p < 0.01; Aspartame VS Sucrose: Z = 5.195, p < 0.001; Sucrose VS Water: Z = -4.442, p < 0.001 ; Sucrose VS Citric Acid: Z = -5.520, p < 0.001; Sucrose VS Caffeine: Z = -5.489, p < 0.001; Citric Acid VS Caffeine: Z = -2.154, p < 0.05; Citric Acid VS Water: Z = -5.073, p < 0.001; Caffeine VS Water: Z = 4.782, p < 0.001; Wilcoxon) (Figure 8f). In general, tongue protrusion gapes occurred the most during sucrose ingestions, followed by control and then aspartame ingestions. Tongue protrusion gapes occurred less often during caffeine and citric acid ingestions.
Regarding mouth gapes, only sucrose and caffeine differed significantly in the frequency of this behavior (Z = -2.271, p < 0.05, Wilcoxon). No other significant differences were found in the frequency of mouth gapes across substances (Aspartame VS Caffeine: Z = -1.134, p > 0.05; Aspartame VS Citric Acid: Z = -0.175, p > 0.05; Aspartame VS Water: Z = 0.333, p > 0.05; Aspartame VS Sucrose: Z = 1.095, p > 0.05; Sucrose VS Water: Z = -1.027, p > 0.05; Sucrose VS Citric Acid: Z = -1.098, p > 0.05; Citric Acid VS Caffeine: Z = 0.707, p > 0.05; Citric Acid VS Water: Z = -0.272, p > 0.05; Caffeine VS Water: Z = 1.190, p > 0.05, Wilcoxon).
The frequency of lip stretching differed significantly between all substances (Aspartame VS Caffeine: Z = 3.195, p < 0.01; Aspartame VS Sucrose: Z = -2.484, p < 0.05; Sucrose VS Citric Acid: Z = 2.673, p < 0.01; Sucrose VS Caffeine: Z = 4.135, p < 0.001; Citric Acid VS Caffeine: Z = -3.624, p < 0.001; Caffeine VS Water: Z = -3.996, p < 0.001; Wilcoxon) with the exceptions of the pairwise comparisons between aspartame and citric acid (Z = -0.034, p > 0.05, Wilcoxon). Trends rather than significant differences were found for the pairwise comparisons between aspartame and tap water (Z = -1.807, p = 0.071, Wilcoxon), citric acid and tap water (Z = 1.941, p = 0.052, Wilcoxon), and sucrose and tap water (Z = 1.732, p = 0.083, Wilcoxon) (Figure 8g). In general, lip stretching occurred the most during caffeine ingestions, followed by citric acid ingestions.
The frequency of lip smacking differed significantly between aspartame and caffeine (Z = -2.085, p < 0.05, Wilcoxon), aspartame and sucrose (Z = 1.968, p < 0.05, Wilcoxon), sucrose and caffeine (Z = -2.612, p < 0.01, Wilcoxon), and sucrose and citric acid (Z = -2.615, p < 0.01, Wilcoxon) (Figure 8h). Trends rather than significant differences were found for the pairwise comparisons between aspartame and citric acid (Z = -1.868, p = 0.062, Wilcoxon), caffeine and tap water (Z =1.899, p = 0.058, Wilcoxon), and citric acid and tap water (Z = -1.657, p =
29
0.098, Wilcoxon). No significant differences were found for the pairwise comparisons between aspartame and water (Z = 0.422, p > 0.05, Wilcoxon), caffeine and citric acid (Z = 0.846, p > 0.05, Wilcoxon), and sucrose and water (Z = -1.186, p > 0.05, Wilcoxon). In general, lip smacking occurred most often during sucrose trials, followed by control trials and then aspartame trials. Lip smacking occurred least often during citric acid trials, followed by caffeine trials.
No significant difference was found between the frequencies of nose wrinkles across substances (Aspartame VS Caffeine: Z = -1, p > 0.05; Aspartame VS Citric Acid: Z < 0.001, p > 0.05; Aspartame VS Water: Z = 0.816, p > 0.05; Aspartame VS Sucrose: Z = -1, p > 0.05; Sucrose VS Water: Z = 1.342, p > 0.05; Sucrose VS Citric Acid: Z = 1, p > 0.05; Sucrose VS Caffeine: Z = -1, p > 0.05; Citric Acid VS Caffeine: Z = 1, p > 0.05; Citric Acid VS Water: Z = -0.816, p > 0.05; Caffeine VS Water: Z = 1.342, p > 0.05, Wilcoxon).
Regarding withdrawals from the dropper, frequencies differed significantly between all tested substances (Aspartame VS Citric Acid: Z = 2.092, p < 0.05; Aspartame VS Sucrose: Z = -3.804, p < 0.001; Sucrose VS Water: Z = 3.140, p < 0.01; Sucrose VS Citric Acid: Z = 4.964, p < 0.001; Sucrose VS Caffeine: Z = 4.696, p < 0.001; Citric Acid VS Water: Z = 3.634, p < 0.001; Caffeine VS Water: Z = -2.938, p < 0.01; Wilcoxon), with the exception of the pairwise comparisons between caffeine and citric (Z = 0.863, p > 0.05, Wilcoxon) and aspartame and tap water (Z = -1.370, p > 0.05, Wilcoxon). The pairwise comparison between aspartame and caffeine revealed a trend rather than a significant difference (Z = 1.900, p = 0.057, Wilcoxon). In general, monkeys withdrew from dropper most often during citric acid, caffeine and aspartame trials, and withdrew the least during sucrose and control trials.
30 * * * * * * * * * * * * * * * * a) b) c) d) e) f) g) h) * *
31
Figure 8: Behavioral frequencies recorded for the group of four spider monkeys for a) licking, b) sniffing, c) eyes open less than 50%, d) eyes closed, e) flat tongue protrusions, f) tongue protrusion gapes, g) lip stretching and h) lip smacking when presented with different taste stimuli. The numbers 1 through 5 correspond respectively to 1 – aspartame, 2 – caffeine, 3 – citric acid, 4 – tap water and 5 – sucrose. The brackets marked with an asterisk (*) represent significant differences in behavioral frequency between substances (p<0.05). The boxes represent the distribution of behavioral frequencies, with the median value represented by the horizontal lines in bold. The bottom and top whiskers of each box represent, respectively, the
lowest and highest recorded frequency for the behavior. The circles (๐) represent the suspected
outliers and the asterisks which are not on top of brackets represent the outliers.
5. Discussion
The spider monkeys’ responsiveness to the different taste substances varied considerably. Only six of the ten substances tasting sweet to humans were preferred over water by all individuals. With four of the ten substances only one or two of the animals were able to detect their sweetness and at least two of the monkeys rejected them.
5.1. Within-species comparison of taste preference thresholds between sweeteners
The spider monkeys’ taste preference thresholds are displayed in Figure 9, alongside with those of chimpanzees and humans (see section 5.2. Comparison of ranking of sweetening potency order between spider monkeys, chimpanzees and humans). The lowest taste preference threshold, and thus the highest taste sensitivity, was recorded for the sweet-tasting protein monellin. However, this value was recorded for only one of the four monkeys. The second and third lowest thresholds correspond to alitame and sodium saccharine, respectively, revealing that the monkeys are more sensitive to these substances than to the remainder of the sweet-tasting substances tested in the present study. The fourth lowest taste preference threshold was the one recorded for thaumatin, also displayed by only one of the four monkeys. Following thaumatin, acesulfame K had the fifth lowest taste preference threshold recorded. As alitame, sodium saccharine and acesulfame K were perceived by all spider monkeys and as the animals displayed taste preference thresholds at least one factor of 10 lower than that of sucrose, these sweet-tasting substances can therefore be described as high-potency sweeteners in this species.
32
The spider monkeys displayed the highest taste preference threshold, and thus the lowest sensitivity, for sorbitol, showing that they are less sensitive to this substance than to the other sweet-tasting substances tested in the current study. Aspartame and isomalt, in order of mention, had the highest taste preference thresholds following sorbitol. As the taste preference thresholds for these three substances are higher than that of sucrose, they can be described as low-potency sweeteners for the spider monkeys.
In the context of discussing the obtained taste preference thresholds of the spider monkeys, it is important to mention any methodological limitations that may have affected the results. In particular, it is relevant to ponder upon the Clever-Hans effect, since the experimenter presenting the bottles to the animals was visible to them during all trials of the two-bottle preference tests. This raises the possibility that the animals may have picked up on any unintended cues that the experimenter may have unconsciously given and thus potentially biasing the animals’ choice of which bottle to preferentially drink from. Although it is unlikely that any potential cues could have been so systematically transmitted from experimenter to monkey to the extent of biasing the obtained results (which require tens of trials for each one of the tested substances), it is not possible to completely rule out this possibility. Nonetheless, the fact that I underwent a training period during which my supervisors observed my posture and behavior when testing the animals and corrected any potential cues, increases my confidence in my results. Additionally, several trials were performed not by me (the main experimenter) but by other experimenters, with no noticeable differences in the animals’ consumptive behavior, further supporting the results.
33
Figure 9: Taste preference thresholds of spider monkeys ( ), chimpanzees ( ), and humans
( ) to sorbitol, aspartame, isomalt, galactose, cyclamate, acesulfame K, thaumatin,
saccharin, alitame and monellin. For spider monkeys ( ), the substances appear in ascending order of sensitivity (corresponding to a descending order of taste preference threshold values). The brackets mark the substances for which only one or two of the monkeys displayed a preference.
34
5.2. Comparison of sweetening potency order between spider monkeys, chimpanzees and humans
Figure 9 (section 5.1. Within-species comparison of taste preference thresholds between sweeteners) displays the available taste thresholds for the tested taste substances in the present study for chimpanzees and humans. Similar to the findings reported here for spider monkeys, alitame, sodium saccharin and acesulfame K also qualify as high-potency sweeteners in chimpanzees (Henderson 2019). However, in contrast to spider monkeys, aspartame is also described as a high-potency sweetener in chimpanzees (Henderson 2019). For both spider monkeys and chimpanzees, sorbitol is described as a low-potency sweetener. Despite the similarities between the ranking order of sweetening potencies of spider monkeys and chimpanzees, no significant correlation was found when comparing the two (rs = 0.617, p > 0.05). Note that isomalt and sodium cyclamate were not included in the statistical analysis for this comparison, as there are no available taste preference threshold values for these substances in chimpanzees.
Of the substances described as high-potency sweeteners for spider monkeys, only sodium saccharin holds the same description for humans (van Gemert 2011). Similar to spider monkeys, sorbitol is described as a low-potency sweetener in humans (van Gemert 2011). A comparison between the ranking order of sweetening potency of spider monkeys and humans (see Figure 9) showed a significant correlation between the order of ranking for the two species (rs = 0.782, p < 0.01). Please note that the higher ranking of sucrose for humans is explained by the difference between thresholds reported in different studies, probably reflecting differences in method and threshold criterion.
For both comparisons it is relevant to remark that for two of the higher-ranked substances in terms of sweetening potency for spider monkeys, monellin and thaumatin, a threshold was only determined for one of the individuals. This may limit our ability to extend the conclusions taken from these results to the population level.
The same comparison of ranking order of sweetening potencies for a similar set of substances to the one used in the present study between chimpanzees and humans was reported by Henderson (2019), showing a higher degree of similarity between the rankings of these two catarrhine primate species than between spider monkeys, which are platyrrhine primates, and chimpanzees and humans, respectively. Given the taxonomic position of these three species within the order of primates, these comparisons suggest that phylogenetic relatedness may
35
affect the degree of similarity or dissimilarity in the relative order of sensitivity for sweet-tasting substances.
5.3. Between-species comparisons of taste preference thresholds
5.3.1. Artificial sweeteners
Table 2 shows a compilation of taste threshold values for the artificial sweeteners acesulfame K, alitame, aspartame, isomalt, sodium cyclamate and sodium saccharin in different species. The spider monkeys’ taste preference threshold for acesulfame K ranges between 0.2 and 1 mM, which is approximately 3 to 25 times higher than that of humans (0.04-0.07 mM) (van Gemert 2011). However, it should be noted that the procedures used to determine taste thresholds in humans allow for the determination of taste detection thresholds rather than taste
preference thresholds. Thus, the taste thresholds for humans represent the lowest concentration
of a given substance that human subjects are able to detect but not necessarily prefer over water. As it is possible that animals detect lower concentrations of a given taste substance than the lowest one preferred over water (Spector 2003), it is then plausible to assume that spider monkeys may be able to detect acesulfame k concentrations as low as humans do.
For chimpanzees, the reported taste preference threshold for acesulfame K is 0.5-2 mM (Henderson 2019), which overlaps with that of spider monkeys. To the extent of the information I gathered, there is no reported threshold for acesulfame K for any other catarrhine primate species.
Within the Platyrrhini, in addition to spider monkeys, the common marmoset (Callithrix
jacchus) was tested for its responsiveness to acesulfame K. However, only a single
concentration (13.8 mM) was tested which was preferred over water (Danilova & Hellekant 2004). Further tests would be needed to clarify if this species prefers acesulfame K at concentrations as low as the ones here reported for spider monkeys.
Within the Strepsirrhini, only the grey mouse lemur (Microcebus murinus) has been included in gustatory responsiveness tests with acesulfame K. However, only a single concentration was tested (7.3 mM), which did not elicit a preference for the substance (Schilling et al. 2004). Electrophysiological studies reported responses in nerve fibres triggered by millimolar concentrations of acesulfame K in the grey mouse lemur (Hellekant et al. 1993). However, it is not certain that such responses signify that the animals would prefer such concentrations
36
over water. A hypothetical generalized lack of responsiveness to acesulfame K in the grey mouse lemur could potentially be explained by its highly plastic generalist diet, in the sense that no specialization towards sweet-tasting food items such as fruits is displayed by this species (Dammhahn & Kappeler 2008).
Non-primate species whose taste responsiveness to acesulfame K have been tested include cows (Bos taurus) and mice (Mus musculus) with reported taste preference threshold ranges of 1-2 mM and 0.3-10 mM, respectively (Hellekant et al. 1994, Bachmanov et al. 2001). Hamsters (Mesocricetus auratus) were reported to prefer a single concentration (4 mM) over water, but no threshold was determined (Danilova et al. 1998). Pigs (Sus scrofa domesticus) displayed a taste preference threshold for acesulfame K of 0.24-1.73 mM (Glaser et al. 2000), which is similar to the one determined for spider monkeys (0.2-1 mM). Thus, the spider monkeys’ sensitivity to acesulfame K appears to be similar to that of non-primate species. This finding is difficult to interpret given the distinct diets and relatively distant taxonomic classifications of the above-mentioned species.
Table 2. Taste preference thresholds (in mM) for the artificial sweeteners acesulfame K, alitame, aspartame, isomalt, sodium cyclamate and sodium saccharin in primates and non-primate mammals.
Species Acesulfame
K
Alitame Aspartame Isomalt Sodium
cyclamate Sodium saccharin Ref. Hominoid primates Pan troglodytes verus 0.5 – 2 0.5 – 2.5 0.5 0.2 – 2 [1] Homo sapiens 0.04 – 0.07 0.0015625 – 0.0125 0.02 – 0.2 5 – 40 0.28 0.00003 – 0.03 [2,3] Catarrhine primates Cercocebus atys atys 0.6 [4] Cercopithecus nictitans 0.5 [4] Macaca fuscata 0.3 [5] Macaca mulatta 0.5 [4] Platyrrhine primates Ateles geoffroyi 0.2 -1 0.002 – 0.005 20 10 – 20 1 0.002 – 0.5 [3] Callithrix jacchus
13.8* 0.3* No pref. No pref. No pref. [6]
Strepsirrhine primates
37 Eulemur mongoz No pref. [7] Microcebus murinus
No pref. No pref. No pref. No pref. No pref. [8]
Varecia variegate
No pref. [9] Non-primates
Bos Taurus 1 – 2 No pref. 4 [10]
Mesocricetus auratus
4* No pref. No pref. No pref. 1.6* [11]
Mus musculus 0.3 – 10 No pref. No pref. 0.43 - 4.3* [12]
Ovis aries No pref. [13]
Rattus norvegicus 0.08 [14] Sus scrofa domesticus 0.24 - 1.73 0.3 No pref. No pref. 2.18 - 4.36 [15]
Please note that the values presented for Homo sapiens are taste detection thresholds. Values marked with one asterisk indicate that a preference was shown for the corresponding concentration, but no taste preference threshold has been determined for that species and substance. “No pref.” indicates that no preference was shown for the substance over water in two-bottle preference tests.
[1] Henderson (2019); [2] van Gemert (2011); [3] Present study; [4] Glaser (1992); [5] Sato et al. (1977); [6] Danilova & Hellekant (2004); [7] Hellekant et al. (1993); [8] Schilling et al. (2004); [9] Nicklasson (2015); [10] Hellekant et al. (1994); [11] Danilova et al. (1998); [12] Bachmanov et al. (2001); [13] Goatcher & Church (1970); [14] Stumphauzer & Williams (1969); [15] Glaser et al. (2000).
The spider monkeys’ taste preference threshold for alitame ranged from 0.002 to 0.005 mM. These threshold values are 600-1500 times lower than those for sucrose in spider monkeys, making alitame a high-potency sweetener in this species. The human threshold for alitame ranges between 0.0015625-0.0125 mM, which overlaps with that of spider monkeys. As it is possible that spider monkeys may detect lower concentrations of a given substance than the one determined as their taste preference threshold, it is plausible to assume that spider monkeys may have a similar or even higher sensitivity to alitame compared to humans. For chimpanzees, the taste preference threshold has been reported at 1 mM (Henderson 2019), at least 200 times higher than that of spider monkeys. This suggests a higher sensitivity to alitame in spider monkeys than in chimpanzees, which could be explained in terms of degree of frugivory. Spider monkeys are highly specialized on the consumption of ripe fruits, and have been suggested to use sweetness as a criterion for food selection (Di Fiore et al 2008; Gonzalez-Zamora et al. 2009), which could ultimately account for an increased sensitivity to sweet taste.
