Husbandry and methodology for assessing learning and memory in the juvenile life stage of octopus

Abstract

The purpose of this project was to use modified methodology and new approaches to determine whether the California two-spotted octopus (Octopus bimaculoides) shows evidence of higher cognitive function as juveniles. This species’ cognitive ability was assessed in ∼4 month old octopus using a food preference test and a learning test (ability to recognize a habitat created from 3D printed rocks and navigate to its hidden food source). Methods for determining associative learning for this species were developed. In addition, potential enhancements to future O. bimaculoides husbandry and study design are discussed.

• •Enrichment and care methodology for juvenile cephalopods from hatchling to juvenile
• •In the article we discuss new approaches for studying associative learning, spatial learning, and food preference that can be adapted for various species of cephalopods, and the usage of 3D printing as a habitat re-creation tool in aquaria
• •We introduce histology methodology for observing and comparing brain development in juvenile cephalopods

Graphical abstract

Specifications table

Subject area:	Agricultural and Biological Sciences
More specific subject area:	Associative and spatial learning
Name of your method:	Approaches to studying spatial and associative learning in cephalopods (Octopus bimaculoides)
Name and reference of original method:	Boal, J.G., Dunham, A.W, Williams, K.T., Hanlon, R.T. (2000). Experimental Evidence for Spatial Learning in Octopuses (Octopus bimaculoides). Journal of Comparative Psychology. 114 (3), 246-252. doi: http://dx.doi.org/10.1037/0735-7036.114.3.246
Resource availability:	PLA Fiber Fish Safe Epoxy Screen Mesh Plant Decor Full Spectrum Lights Stones for Zones Purit (Activated Carbon)

Method details

Individual husbandry

Samples

We bought twelve Octopus bimaculoides (California Two-spotted) hatchlings (∼6–7 mm in mantle length and 7 days post-hatch) from the Marine Biological Laboratory in Woods Hole, Massachusetts. These hatchlings were delivered fully developed and able to survive on their own [12]. We raised them to ∼25 mm (∼4 months) before experiments began as they begin to forage at this point in their development [10]. Of the twelve individuals, six octopuses died from unknown causes before the start of experiment.

Octopus housing and care

All animals were housed in separate aquaria, with dimensions: 60.96 cm L × 30.48 cm D × 60.96 cm W for two, 60.96 cm L × 31.75 cm D × 60.96 cm W for one, and 46.99 cm L × 29.21 cm D × 62.23 cm W for six. The size of this aquaria should be adjusted to accommodate the size of the cephalopod being utilized to ensure visuality for health checks and experimentation.

All aquaria contained thirty pounds of crushed coral, 1–2 approximately 10 cm in length artificial plants, and a 10 cm long terra clay pot for concealment [18]. Substrate can either be removed before experimentation to prevent excessive concealment, or not removed entirely. The substrate can prevent visualization of smaller species during health checks and feedings. Plants and pots were spread throughout housing aquarium until two weeks before experiments began, and then moved to one side of aquaria.

Mesh lids were velcroed along the sides of aquaria to prevent removal. The water level was kept 5–8 centimeters below the top to prevent escape. Each tank was maintained at 23–25 °C with heaters to encourage faster growth [10]. Since octopus are sensitive to certain compounds such as ammonia (NH₃), nitrite (NO₂), and copper (Cu), water chemistry was checked regularly. Salinity was checked daily to ensure concentrations remained at a specific gravity of 1.026, pH of 8–8.4, NO₃, NO₂, and copper of 0, and NH₃ < 30. The use of a hydrometer and marine chemistry test kits ensured these water quality parameters, and the protein skimmer in the filtration sump assisted in minimizing high concentrations of organic waste, biological toxins, and phosphate. To maintain these water quality parameters, 20% water changes were completed as needed. The water in the system was treated routinely with activated carbon to prevent unwanted olfactory cues between individuals, and nitrifying bacteria (Turbo Start 900 or API Aqua Essential) to balance nitrite and ammonia levels [14]. Two identical systems containing 6 tanks each were maintained. Fig. 1 illustrates the holding system holding five octopuses as one octopus was deceased at this time.

Fig. 1.

Aquaria Filtration System Design. Five of the aquaria were connected by a series of PVC pipes to one sump system (114-227 liters), with a protein skimmer and overflow compartment. Activated carbon was placed into the sump when needed.

Diet

The natural diet for O. bimaculoides in the wild is mollusks and crustaceans. To simulate this natural diet, the octopus were fed a combination of mysid shrimp and amphipods/copepods. Upon arrival to two-three months post-arrival, we decapitated the ghost shrimp and cut it into tiny pieces to be handed directly to the octopus. At that point, octopuses began hiding more, so we began feeding juveniles live food [14]. Food items were immobilized by placing into perforated 15 ml centrifuge tubes that allowed octopus to reach inside. All food items were kept in separate aquaria with their own attachable filters as needed. Brine shrimp and ghost shrimp were fed every other day with TDO ChromaBoost and shrimp pellets respectively. Octopus individuals were fed twice daily until they were more than three months of age [13]. Once they could eat on their own, they were fed once in the afternoon every other day since previous observations noticed eating occurring every other day and a decrease in food consumption to once a day.

Enrichment

Hatchlings (< 2 months) were given enrichment starting four weeks after octopus were purchased. They were given a variety of textured and/or colored objects for one week at a time or novel food items for one day (Table 1). Enrichment is an important aspect of raising species with advanced cognition since it stimulates and encourages natural behaviors. Octopus are curious creatures that enjoy different textures and shapes [7]. Introducing different food types encouraged problem solving and manipulation common in their natural habitats [3]. Even the change over from dead, hand-fed prey to hunting for live items served as enrichment since this required the use of problem solving and dexterity. During the first few months of development, octopus juveniles (> 2 months) preferred areas of concealment, so enrichment options used took this into account [7]

Table 1.

Enrichment Schedule. After four weeks of being in the lab, the individuals were introduced to various enrichment types, except during periods of experiments.

Week	Enrichment Type
Week 1	2 Artificial Goldfish (varying colors)
Week 2	PVC Pipe
Week 3	Artificial Jellyfish
Week 4 – Week 5	Tall/Bushy Artificial Plant
Week 6	2-inch PVC Pipe
Week 7- Week 8	No Enrichment
Week 9	All housing material moved to side under water flow
Week 10	Brine Shrimp introduced
Week 11	Food placed into 15 mL centrifuge tubes
Week 12 - Week 15	No Enrichment: In Experiment
Week 16- Week17	Tall/Bushy Artificial Plant
Week 18–28	No Enrichment: In Experiment

Food preference

The first part of this experiment determined the food preference of O. bimaculoides for future trials. For one week a gram of fresh shrimp, ghost shrimp (the food item used for the juveniles up to this point), crab, and scallops were offered, each in a separate glass jar marked one through four [2,3,11,12,17]. The jars were placed on their sides, openings facing in the same direction and were ∼13 cm from each other (5× the mantle length of the octopus). They formed a diamond in the middle of the aquaria (Fig. 2). Each jar sat in the middle of a ∼7.62 cm diameter ring of black aquarium rocks. Each ring represented a zone (1–4) and was numbered to match the jar with the respective food item. Anything outside the rings was considered the neutral zone. The order of the jars in the diamond positioned zones was randomized for each octopus [15].

Fig. 2.

Configuration of Zones and Rocks for Food Preference Experiment. Ring of black aquarium rocks were considered zones with space outside them being the neutral zone. One glass jar labelled between 1 and 4 was placed in each.

For example: Octopus 1 had the following order from the top of the diamond in the clockwise direction: 4, 3, 2, and 1. Octopus 2 had the following order from the top of the diamond in the clockwise direction: 1, 2, 3, and 4.

The ZOSI cameras were placed over each aquarium with the use of a PVC stand to have all four jars in the field of view when recording the experiment (Fig. 3).

Fig. 3.

Camera Design for Food Preference and Associative Learning Experiments. Cameras were attached with rubber bands to interlocked PVC tubes and placed directly over zones.

The food samples remained in the aquarium for 24-hour sessions every day. The recordings were switched from every other day after two sessions to every day to ensure all activity was captured.

During the sessions, behavioral responses were observed: hiding (i.e. not visible within the aquaria), eating (i.e. actively interacting with food item), exploring (i.e. actively investigating the jars within zones), resting (i.e. inactive but visible), and swimming (i.e. motion without touching the floor).

The time spent doing each activity for each zone was recorded, as well as how many times a switch (i.e. change in activity and/or zone) took place during the session. The amount of time spent with each food item (i.e. time in the food item jar) out of the total time spent with all food items (i.e. total time spent in all jars) was also measured to assist in determining preferred food.

For the second part, the comparison of frozen versus fresh options would have been done for the preferred food option, with three jars containing the least favored food and one random jar containing the favored food choice.

Note: The ZOSI cameras were unable to clearly visualize the octopus's movement with the vast difference in octopus versus aquaria size. Utilizing stronger cameras or ensuring organism and aquaria are proportional is highly recommended.

Associative learning

The ZOSI cameras remained recording for 24 h every other day. One jar was marked with a black x pattern, while the other three jars were marked with the horizontal black lines. The one with the x had the preferred food option (i.e. the food item with the highest percentage of visits in last experiment) and the one with the horizontal lines had the least preferred option (i.e. the food item with the lowest percentage of visits in last experiment). The jars were randomly positioned in the aquaria in each session to ensure learning of pattern and not location. The zones were still used as before, and the same behaviors and variables were observed. The experiment started with four trials (i.e., session every other day as before) and increased as needed for each individual to reach a higher percentage (i.e., time with food item) for the preferred food versus the least preferred food.

The jars were then switched so that the preferred food source was now in the jar with horizontal lines. Time was then recorded for how long it took for the individual to begin associating the new jar with the preferred food choice. Experiments once again began with four trials and increased as needed. Once a higher percentage (i.e., time with food item) was observed, experiment was stopped.

For the last part of the experiment, a 50 ml centrifuge tube with ∼1 inch holes along the sides was partially buried into the substrate and used to hold the preferred food. The tube was marked with the horizontal lines seen on the jar. Three jars were marked with the x pattern but contained the least preferred food source. The position of each jar and the centrifuge tube was randomized each session and still remained in their assigned zones. The same variables and behaviors were observed as previously stated. This experiment would train the octopus to enter a controlled setting where there is live food, allowing for use in future spatial learning test.

Spatial learning

During the experiment octopus were given one brine shrimp inside a 15 ml centrifuge tube for each trial, and one free swimming ghost shrimp on Saturdays. During weeks without experiments, they were given one free swimming ghost shrimp every other day until Sunday.

Experimental design

During the experiment, the housing equipment (pot and plants) were moved to the side of each aquarium closest to the flowing water. Each aquarium had a video camera (ZOSI 1080P 8CH Security Camera System) overhead recording all activity in two-hour increments to a hard drive in the laboratory. The pump from the sump to the aquaria was turned off during experiment to decrease reflection on water surface during recording.

Three-dimensional, identical prints of Texas Holey Rocks were printed with white PLA (polylactic acid) filament in a 3-D printer. They were then glued with fish safe epoxy to a platform, and/or each other, to allow easy removal and addition to aquaria. Each platform had an increasing number of rocks present, beginning with two, three, and then four, to create increasing structural complexity (See Graphical abstract). Rock 1 had the dimensions: 16.51 cm L × 63.5 cm W × 15.24 cm H. Rock 2 had the dimensions: 15.24 cm L × 20.32 cm W × 12.7 cm H. Three-dimensional printed rocks were found to float in saltwater and so were anchored with bricks rubberbanded to the bottom of the platform. The habitat platform was inserted starting ten minutes before experimentation session began to allow for all rock formations to be placed – and adequately immersed – before experiments began.

During experiments, black cardstock paper was attached to the front of the aquaria with Velcro to allow for easy removal during set up. Remaining sides were covered with taped black cardstock paper to prevent distraction from other octopus, outside light sources, and movement. Lamps with 40 watts of red-blue light were placed over aquaria for twelve hours a day, which included the four hours of experimentation. Since these octopus are already known for their exploratory movements, this encouraged their natural habits [4,19].

Experimental aquaria rotation

Octopus were divided into groups with three individuals per group. Due to pre-experiment mortality, only six individuals of the original twelve individuals were left to perform the experiment.

The groups were as follows.

• •Group 1: Individual 1, 4, 5
• •Group 2: Individual 8, 10, 12

Each group were introduced to the experimental habitats in randomized order (Table 2). The subjects were not fed 24 h before each experiment day to encourage foraging [4,20]. Two trials (9–11am and 4:30–6:30pm) were done on experiment days (Monday, Wednesday, and Friday).

Table 2.

Weekly Schedule for Introduction to Different Complexity Tanks. Each group member was tested twice a day for two hours, once in the morning and once in the afternoon. Each individual started at a different level of complexity and rotated through each level over three days as indicated below. Both groups were tested each day.

Group Number	Individual	Complexity Schedule (AM & PM)	Schedule for Experiment
1	Individual 1	low complexity (experimental tank 1) to high complexity (experimental tank 3)	Day 1, 3,5
1	Individual 4	high complexity (experimental tank 3) to medium complexity (experimental tank 2)	Day 1, 3, 5
1	Individual 5	Medium complexity (experimental tank 2) to low complexity (experimental tank 1)	Day 1, 3, 5
2	Individual 8	low complexity (experimental tank 1) to high complexity (experimental tank 3)	Day 1, 3, 5
2	Individual 10	high complexity (experimental tank 3) to medium complexity (experimental tank 2)	Day 1, 3, 5
2	Individual 12	high complexity (experimental tank 3) to medium complexity (experimental tank 2)	Day 1, 3, 5

If no mortality was experienced, the individuals went through an eight-week training period, for a total of 16 training trials per individual, where they were rotated through all three levels of experimental habitats for a total of six trials a week with two trials for each of the three habitats (See Graphical abstract; [1,4,6,9]).

For Example: Individual 1 (Group 1) for the trials had Habitat 1 (low complexity) on Day 1, Habitat 2 (medium complexity) on Day 3, and Habitat 3 (high complexity) on Day 5.

They then spent one week off before being introduced to the aquaria again. They would take an additional four weeks off and then be tested again for one trial (Tables 2 & 3). The training period was used to examine learning, while the extended period looked at memory.

Table 3.

AM & PM Experimental Aquaria Introduction Schedule. This chart shows what level each individual in each group is introduced to for every day until end of experiment. Each color indicates a specific individual. A gantt chart was used to create this diagram.

Statistical models for experiment I

Linear Mixed Effects Model: This experiment was attempting to determine the effect of two explanatory variables: 1) “time” (i.e., time away from habitat; categorical with 2 levels: 1wk and 1mth) and 2) “complexity” (i.e., habitat complexity; categorical with three levels; low, medium, and high) on “pathway length” (i.e., distance traveled prior to finding food, which is indicative of activity) and “time to find food” (i.e., total time taken for octopus to find the food item once added to the experimental arena). Note that learning would lead to a shorter pathway length to food source and time to find food as trials progress. There were also three individuals participating at once in each “round” of trials. To account for the repeated measure within individuals and the potential for greater variation within rounds than among round, each trial included a random effect in which individual was nested within trial. Models are outlined below.

• 1.PathwayLength (activity) ∼ Time*Complexity + (1|Trial/Individual)
• 2.TimeToFindFood (learning) ∼ Time*Complexity + (1|Trial/Individual)

Brain sectioning

The whole octopus brain (n=4) was imaged using histology techniques with the following ages: 3 weeks, 8 weeks, 11 weeks, and 13 weeks. First, the tissues were decalcified in Formical-4 for 24 h to remove the beak and radula. The octopus were then rinsed with distilled water and stored in 70% ethanol (ETOH).

For paraffin embedding, samples were placed in microcentrifuge tubes with 2 mL of the following solution (each solution was removed with a pipette between each step) for these specific periods of time: 80% ETOH for 10 min, 90% ETOH for 15 min twice, 100% ETOH for 20 min thrice, and finally xylene for 25 min four times. Then, 2 mL of paraffin was added, and the tissues were incubated overnight at 56 °C. The following day the xylene/paraffin mixture was changed, and incubation was continued at 56 °C. The changing of the xylene/paraffin mixture was repeated twice, once per hour. The entire octopus was then moved to a mold of fresh, melted paraffin. Each octopus was allowed to cool overnight. Paraffin embedded samples were cut into longitudinal sections in 10 µm layers.

These sections were then stained in a 12-compartment staining rack. The compartment was filled to 150 mL to the top of the unfrosted portion of the slides. In compartment 1 and 2 the slides were deparaffinized in xylene for 10 min. In compartment 3, the slide was rehydrated in 100% ETOH for 10 min. In compartment 5 the slide was rehydrated in 95% ETOH for 2 min (150 mL = 143 mL 95% ETOH + 7 mL DI H₂O). In compartment 6, the slides were rehydrated in 70% ETOH for 2 min (150 mL = 105 mL 95% ETOH + 45mL DI H₂O). The slides were washed dipped 10 times in distilled water. In compartment 7, the slides were stained in Harris’ hematoxylin for 2 min. The slides were washed in running tap water for 5 min. In compartment 8, differentiation in 1% of acid alcohol was done for 30 s. Slides were washed again in running tap water for 3 min. In compartment 9, the samples were blued in saturated lithium carbonate for 30 s and the washed in running water for 5 min. Compartment 5 was used to rinse the slides 10× in 95% ETOH. Compartment 10 was used to counterstain the slides in Eosin Y for 60 s. Compartment 5 was used to dehydrate the slides in 95% ETOH for 5 min. Compartments 3 was then used to dehydrate the slides in 100% ETOH for 10 min. Compartments 11 and 12 were used to clear slides with xylene for 5 min. The coverslips were mounted with Permount before the xylene dried and the slides were allowed to dry for 24 h. The slides were then visualized with a compound microscope under 50× and 100× and the brain sizes of octopus at each age (3–13 weeks) was measured with the use of the LAS X program. Area of the sample visible on the slide was measured as well as the area of the brain. The length and width of the brain was measured as well.

Method validation

Organisms for this experiment were deceased before completion of experiments, but successful tests for long term and short term memory done by other researchers utilized a similar design and schedule method utilized here [4,5]. Methodology used for brain histology were compared to those done for Octopus vulgaris (common octopus), as well as older Octopus bimaculoides and were found to reveal similar results [8,16].

Ethics statements

The protocol was reviewed and approved by the NSU Institutional Animal Care and Use Committee (IACUC), as of July 6th, 2021 with protocol number 2021.07.TDS1. These experiments complied with with the ARRIVE guidelines and were carried out in accordance with the U.K. Animals (Scientific Procedures) Act, 1986 and associated guidelines; EU Directive 2010/63/EU for animal experiments; or the National Institutes of Health guide for the care and use of laboratory animals.

Note: Since all octopuses were hatchlings up to juvenile age, sex was not determined and did not influence results of the study.

CRediT authorship contribution statement

Shaquilla Hamlett: Conceptualization, Methodology, Formal analysis, Investigation, Resources, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

• Data will be made available on request.