Spin Side_crop.jpg

Kealakekua Bay Data

Here we explore first how spinner dolphins used the bay and then how humans used the bay. As the human use of the bay was affected by whether the dolphins were present in the bay, or absent, human use for these two states was compared. 

The stars (*) in the figures below indicate statistical significance levels between the No Dolphin (dolphins absent from the bay) and Dolphins Present bars for each zone:

*   - P<0.05

** - P<0.01

***- P<0.001

N.S. - No significant difference.

For more information on the statistical methods used to analyse the data see the methods section in the Final Report.

Screenshot 2021-10-15 at 10.23.01.png

Figure 1.  Kealakekua Bay, with the spinner dolphin core resting area indicated with a red oval. The observation station was located near the Wharf, from where the kayaks were launched. There is also a beach in the North corner of zone C, from where swimmers entered the water. The double headed arrow spans 250m, for scale.

Dolphin Use of Kealakekua Bay

On average, dolphins were sighted in the bay on 45.9% of monitoring days from March 11, 2006 – July 21, 2008, excluding the complete closure period.  When spinner dolphins were in Kealakekua Bay, they spent the majority of their time within zone (B), designed to encompass most of the Core Resting Area.  Dolphin zone use was recorded during activity scans.  However, since the activity scans were 5 minutes long, the dolphins could spend time in more than one zone during a scan.  Thus, the total (2006-08) zone use percentage for the bay (107%, n=334) exceeded 100%, with the dolphins recorded in zone “B” during 73.6% of activity scans (Figure 2).

Screenshot 2021-09-16 at 10.58.13.png

Figure 2.  Zone use by spinner dolphin schools in Kealakekua Bay, as measured during the activity scans.  As each scan was five minutes long, the schools could spend time in more than one zone.  The total percentage therefore exceeds 100%.


Human Use of Kealakekua Bay

Use by Zone

The human use can be divided into three components: (1) swimmers with snorkeling gear; (2) kayakers and an occasional Hawaiian canoe, and; (3) visitors brought to the coral reef in north end of the bay, near the Captain Cook monument, by tour boats from harbors north of the bay, including an 18m long catamaran that moored in the NW corner of zone “A”,  bringing about 100 people per visit, as well as several 6-8m long rigid-hull inflatable boats (RHIB) bringing from six to 20 people and drifting along the NW coastline of zones “A” and “D”.  The swimmers from these boats were not counted when in those zones, but all vessels, including kayaks, were counted and are part of the reported numbers.  In total, 389 snapshots were collected in Kealakekua Bay when spinner dolphins were absent and 210 when they were present.  


Both swimmers and kayakers were more likely to use the three southern zones (B, C, F) in Kealakekua Bay when spinner dolphins were present.  Furthermore, there were 5 times more swimmers and more than double the numbers of kayaks in zone “B” when dolphins were present than when absent (Figure 3a. and 3b.). The increase can be attributed to both swimmers and kayaks moving into zone “B” from adjacent zones, with a significant decrease of swimmers in zone “F” and a significant decrease of kayaks in zones “D”, “E” and “F” (Figure 3).  The increase in the number of swimmers in zone “C”, including the waters by the beach, coincided with observers reporting more swimmers entering the water from the beach into that zone when dolphins were present. 


A Hawaiian canoe was recorded in Kealakekua Bay on 28 occasions, seven of which when dolphins were present.  It was never recorded in zone “B” when dolphins were present (Figure 3b).  


Both small and large motorboats changed how they used the bay, depending on whether the dolphins were present in the bay or not. They were mostly recorded in the two zones (“A” and “D”) near the Captain Cook Monument in the northern part of the bay, 80.5% and 87.5% respectively (Figure 3c.).  However, the number of motorboats of both size classes were significantly higher in zone “C”, and the number of smaller motorboats was lower in zone “D”, when dolphins were present.  Combining all motorboat data provided a significantly higher number of vessels in zone “B” when dolphins were present (P<0.05).

Screenshot 2021-10-15 at 11.50.18.png
Screenshot 2021-10-15 at 11.50.35.png
Screenshot 2021-10-15 at 11.51.00.png

Figure 3a. Use of Kealakekua Bay by swimmers with snorkeling gear, entering the Bay from Napoʻopoʻo beach on the south shore of the bay.

Figure 3b. Use of Kealakekua Bay by kayaks and an Hawaiian canoe entering the bay from the wharf at Napoʻopoʻo on the south shore of the bay.

Figure 3c. Use of Kealakekua Bay by small (≤7.5m) and large (8m) Motor Boats.


Human Use of Kealakekua Bay

Use of the Dolphin Core Resting Area by Time of Day

Screenshot 2021-10-15 at 12.03.56.png
Screenshot 2021-10-15 at 12.04.20.png

Figure 4. Human use of zone “B” by hour in Kealakekua Bay comparing snapshots when spinner dolphins were present versus absent in the bay for a) swimmers and b) kayaks. Significant differences between bars during the same hour are indicated using standard statistical notation.  See methods section for further information.

A further analysis of the human use of the Spinner Dolphin Core Resting Area (zone “B”) by time off day shows that the number of swimmers and kayaks recorded in this zone when dolphins were present was highest during the morning hours. 

The presence of dolphins between 0800-0859 correlated with a 7 times higher swimmer count and 11 times higher kayak count in the Core Resting Area (Figure 4) and a 3 times higher count of both swimmers and kayaks during the 0900 hour.  There continued to be between 2 and 4 times as many kayaks in the Core Resting Area when dolphins were present throughout the day (Figure 4b).  During the afternoon (1200 – 1659), few snapshots were made when swimmers were in the water and no significant differences were found in swimmer use. However, when combining the data for all afternoon time periods for swimmers in the Core Resting Area it showed a significant increase when dolphins were present (P=0.037). 


When dolphins were absent, few swimmers were counted in the Core Resting Area, with the exception of the 0800 hour and 0900 hour, when an average 1.8 swimmers and 2.4 swimmers were counted per snapshot respectively.  For the rest of the day one swimmer was recorded in this zone in less than every third snapshot.  The number of kayaks counted in the Core Resting Area stayed at about 1 per snapshot throughout the day from 0900 to 1659.

Human Distribution Relative to the Dolphins

Monitoring data were also collected on human distribution relative to the dolphins themselves.  Swimmers and kayakers were most frequently recorded close to the dolphins (Figure 5).  On average, 1.9 swimmers were ≤10m of the dolphins during each snapshot, actively pursuing and attempting to interact with them, with another 1.7 swimmers 10m-50m from the dolphins and 0.8 swimmers 50m–200m away.  Thus swimmers observed within 200m of the dolphins were significantly more likely to be closer than 50m from the dolphins than expected (P< 0.001).  An average 0.6 kayaks were <10m from the dolphins, with an additional 1.4 kayaks 10m-50m away and 0.9 kayaks 50m–200m away.  Thus kayaks within 200m of the dolphins were also significantly more likely to be closer than 50m compared to what would be expected (P<0.001).  Thus, on average, 3.6 swimmers and 2.0 kayaks were ≤50m of the spinner dolphins during each of 171 activity-scans, with additional swimmers and kayaks within 200m.

Screenshot 2021-10-15 at 12.08.32.png

Figure 5.  Distribution of human user groups relative to the dolphins during 5-min activity scans (n=171) done between 2006 and 2008.


Adult Aerial Behavior Categories relative to:

Human Proximity

Screenshot 2021-10-15 at 10.49.26.png

Figure 6. Average aerial behavior counts during two conditions, people and vessels over (>) 200m from the dolphins vs. people and vessels within (≤) 100m of the dolphins.  The porpoise only bar (20.4) extends well beyond the graph. Significant differences between bars for the same behavior type are indicated using standard statistical notation. 

Screenshot 2021-10-15 at 10.48.28.png

To assess whether the acrobatic and two types of fast-swim  (non-porpoising and porpoising) behaviors were affected by the human presence, behavior frequencies for two human-dolphin distance categories (≤100m, >200m) were compared.  The sample sizes for the two categories were 88 and 22 respectively.


The mean number of acrobatic and fast-swim behaviors were all significantly higher when humans were ≤100m from the dolphins, while the number of slaps/splashes remained statistically unchanged (Figure 6).  The mean number of behaviors per activity scan was 3-times higher for acrobatic behavior and 2.5 times higher for non-porpoising fast swim behaviors when people were ≤100m from the dolphins compared to >200m.  All porpoising was recorded with people ≤100m.

Time of Day

The pattern of dolphins having significantly higher aerial behavior rates when people were within 100m compared to over 200m away from the dolphins was observed during all time periods except early afternoon, 1200-1400 hours (Figure 7).

Figure 7. Comparing the mean number of aerial behavior when people were within (≤) 100m of the dolphins, versus over (>) 200m from the dolphins during different times of the day.  Significant differences between bars for the same behavior type are indicated using standard statistical notation. 

LeapingCalf Crop.jpg

Aerial Behavior by Calves

Most (33 of 64) of the calf aerial behavior recorded in Kealakekua Bay between Mar 2006 and May 2007 were recorded during the complete closure period, including 50% of the spins and 100% of the spin-flips and fluke slaps. Thus 51.6% of the calf aerials were recorded in the twelve activity-scans taken during the complete closure period, representing only 9.9% of all activity scans. This represents an average 2.8 calf aerials/scan compared to 0.3 calf aerials/scan for the rest of the study. Therefore, the number of recorded calf aerial behavior was more than nine times higher than expected during the complete closure time, compared to the rest of the monitoring period, a highly significant difference (P= 7 x 10    ).



The Human Exclusion Experiment

In the middle of the study, a large earthquake created a landslide in Kealakekua Bay that caused State authorities to completely close down the bay for all human activity for 18 days. Several structures in the nearby village of Napo'o po'o were damaged and the volunteer observers had to abandon their regular observations and deal with the havoc that had been caused to their community. However, one of the volunteers managed to go down to the bay on three days to make his observations.

As a result, this serendipitous natural experiment provided some valuable data on how the dolphins behaved when the humans were absent.

During the complete closure period following the earthquake on October 15, 2006 monitoring was conducted on three days (October 19, 21 and 26), thus providing too small of a sample size for statistical comparisons with the time periods before and after. However, spinner dolphins were present in the bay on all three days, for an occupancy rate of 100%, compared to 39.3% before the earthquake and 31.8% between the time people were allowed back into the area again and April 12, 2007.

Twelve activity scans were conducted over the three monitoring days during the complete closure time, all between 10:14 and 11:45. These activity scans can therefore best be compared to the scans conducted during the 0900-1159 hours before the earthquake and after the complete closure period (Fig. 8).


The dolphins’ behavior changed during the human exclusion period after the earthquake, with a significant reduction in both acrobatic behavior and slaps and splashes. The sample size was not adequate for a statistical analysis of the fast swimming and maneuvering behavior, although the smallest mean (0.08 – 1 record in 12 scans) was recorded during the complete closure period.

After the closure area was reduced from encompassing approximately 56% of zone “B” to about 26%, the frequency of acrobatic behavior increased to once again be significantly higher than during the complete closure. The frequency of slaps and splashes on the other hand did not change significantly (P = 0.33). It should also be noted that the total proportion of all acrobatic and fast swim behavior was only 27% during the complete closure period compared to 53% before and 88% after.

EQ Specifics
Screenshot 2021-10-04 at 14.09.16.png

Figure 8. Mean number of aerial behavior during the 0900-1100 hours in Kealakekua Bay before, during and after the complete closure period. Significant differences between bars for the same behavior type are indicated using standard statistical notation.