I have been following the fates of the wolves in the Voyageurs National Park and Voyageurs Ecosystem.
To my impression, there has been very high morality due to starvation both in pups and adults and the pup survival rate has been rather low.
I attempted to read through the VNP wolf populations reports of 1987 – 1991 (Gogan, P.J. et al., 2004), 2020 – 2021 (Gable, T. et al., 2021) and 2021 – 2022 (Gable, T. et al., 2022) in order to follow these trends and to try and understand the possible causes behind them.
Unfortunately, no access to the reports between 1991 and 2020 has been granted on the internet.
The following estimates have been listed in the VWP Population Report of 2021 – 2022:
There have been several changes – a slight decrease in the main prey population (white-tailed deer) (from 2 – 4.6 deer / km2, 1987 – 2001, to 2 – 3 deer / km2 during the decade leading to 2022).
Beaver, the second prey, densities have been relatively stable although reported as slightly decreasing since 1990-ies.
Other prey (moose, smaller mammals, fish etc.) are taken too rarely to notably affect wolf population trends.
While it has been estimated that wolf densities have grown since the 1980-ies, the current Voyageurs Wolf Project team also recognizes that this could be due to perfected tracking methods.
Namely, the population density has not grown as much as the efficiency in detecting wolves has improved.
However, it has been noticed that pack territory size has decreased which is strange because this should happen upon increase and not decrease in prey populations.
Average pack size has also dropped from 5.5 wolves per pack in 1987 – 1991 to 4.6. wolves per pack in 2021 – 2022.
It is difficult to tell if number of packs has increased because the packs have been replaced by new packs and the 1987 – 1991 pack summaries provide little insight into how many packs there were on the same area as studied in 2021 – 2022.
The earlier report (Figure 5) indicates of the existence of, e.g., 13 – 14 packs in 1988; 11 packs in 1989; 10 packs in 1990 but this only refers to packs with radio-collared members and also it is not clear (to me) how these packs overlap with the study area of the later reports.
For example, the report of 2021 – 2022 discusses the existence of 16 wolf packs but it appears that several packs that used to be studied formerly and that ranged somewhat outside of VNP, have been excluded.
I have come by impression that the number of packs might have grown slightly and that their territories have shrunk – at least the packs below Lake Kabetogama seem to have narrower ranges, tucked more closely together but it is a subjective impression which I cannot quantify with the data that I have.
The pack size seems to have decreased both on Kabetogama Peninsula and below Lake Kabetogama.
The packs on peninsula are still rather large (about 5 ind. in winter counts) albeit it is a drop from earlier data when 7 – 11 wolves were detected.
However, pack size below Lake Kabetogama is now rather small and frequently constitutes 2 – 4 wolves (somewhat larger on the east and the right side and lower in the middle).
It appears to me that pup survival is very low and many pups die even before winter.
The population report of 2021 – 2022, however, ascertains, that pup survival has improved since 2020 – 2021 which is also detected in somewhat larger pack sizes below Lake Kabetogama in winter 2022.
It is very hard to compare mortality causes between the early and the current study periods because of the different tracking methods.
Nevertheless, the 1987 – 1991 report mentions 2 incidents of starvation among radio-collared wolves and 1 incident among non-collared wolves during the study period.
1 of these incidents should be described as starvation due to trauma (a stick wedged between teeth preventing feeding).
This could be reflective, once again, of lower detection rates.
Meanwhile, excluding pup survival (due to any reasons) which was not excluded in the earlier study where some pups were collared and included in mortality analysis, there is 1 incident of starvation among adults in 2021 – 2022 report.
It appears that adult mortality is not the issue and that the adults manage to get by.
However, pup mortality is very high and often pups die even before autumn.
From the reports 2020 – 2022, the following information can be listed (including only packs below Lake Kabetogama because, in my view, that is where the survival rates are lower and result in very small packs that possibly cannot defend territories which is why the territory size shrinks):
Half-Moon Pack (Central):
- By December, 2020 – all pups (4) died;
- By winter 2020/21, all pups (8) died although 7 had been alive until at least August;
- 2021 – 2022, 1 pup survived (litter size unknown).
Huron Pack (Central):
- Female produced pups in spring, 2021 but pack dissolved in August, 2021 possibly because all pups died.
Lightfoot Pack (Central):
- By November 2020, all pups (7) died;
- Lightfoot Pack has had low pup survival rate since 2017.
Moonshadow Pack (Central):
- 2021, 2 pups survived (of unknown litter size);
- Winter 2022, Breeding male died possibly of starvation.
Paradise Pack (Central):
- By July 2020, all pups (4) died;
- Winter 202/21 – 2 out of 5 pups survived.
Tamarack Pack (Left):
- 2021 – 1 pup produced but did not survive.
Whiskey Point Pack (Right):
- By winter 2019/20 – no pups survived;
- Winter 2021 – all (4) pups survived.
Windsong Pack (Central, Lower):
- Winter 2021/22 – 2 pups survived but in May, 2022 were killed for depredation purposes.
Wiyapka Lake Pack (Central but slightly to the Right):
- 2020 – 6 pups produced, all died;
- By winter 2021/22 – pack includes 3 surviving pups.
These pup survival rates might not be abnormal but I see a pattern where the poor pup survival rates negatively affect the pack’s ability to maintain a territory during winter due to a small number of pack members which could result in smaller range, more frequent pack turnovers and intraspecific strife
Intraspecific strife mortality rates do not appear to have increased from 1987 – 1991.
Pack turnover rates cannot be compared between the two studies but in the 2020 – 2022 reports at least 5 incidents of pack displacement or leader replacement (within the pack or by immigrant into the pack) due to several causes have been described.
This could explain the smaller territory size which is characteristic of unstable wolf populations.
However, I believe there is a connection between the low pup survival and the smaller pack sizes not merely due to pack’s inability to maintain ranges without recruited members aiding in territory defense.
I believe there has been some change in prey availability that has caused both low pup survival and the shrinking of the pack sizes.
Even a pair of wolves can maintain relatively large territories if necessary (temporarily).
The inability to defend territories might be linked to modification in landscape parameters.
Perhaps due to the climate change, some areas of water are impossible to be crossed any longer as it used to be possible before (lack of ice cover).
This might naturally delineate wolf territories forcing them into narrower units that do not provide sufficient prey supplies.
Such landscape changes might have affected deer behaviour, as well.
If beaver population has decreased, this could, on other hand, indicate disappearance of streams due to dropping ground water levels, increased summer temperatures or other factors.
In these areas, deer might not choose to raise their young (a significant summer food source that determines pup survival which in VNP appears to be low even during summer and early autumn) due to poor quality or low quantity of vegetation and lack of drinking water.
The deer might be affected both during winter and during summer as they cannot migrate freely if some bodies of water do not freeze over and they cannot find shelter, forage in areas where water regime has been disrupted due to some form of drought.
They might instead gather nearer beaver locations where vegetation is lusher and provides structural resources for sheltering and hiding from predators.
Thus, food resource hotspots might form in certain locations but it depends greatly whether these locations are within the ranges accessible to wolves.
If the ranges have shrunken not due to abundance of food but due to other reasons such as inability do defend territory, population instability and geographic barriers during winter when ranges are often defined, the wolves might have become limited in their access to fawns.
Additionally, the VWP has published several studies on wolf food habits in the VNP.
It appears that during the periods of scarcity in spring/early summer and late summer/autumn, wolves have learned to adapt through utilization of alternative food sources.
E.g., Bowman Bay (later Huron) pack (but also others) specialized on fish during April – May when fawns might not have been born yet but the denning has already begun (Freund, D.R. et al., 2023).
The fact that several of the Central packs below Lake Kabetogama have focused on fish independently one of another suggests a wider scarcity of other food resources.
Wolves also appear to be spending a lot of time near the river during the fish catching season and therefore it seems to be an important food source that effort and time are not spared to acquire (including by breeding males).
Bowman Bay Pack had occupied the same territory from 2016 – 2019 (see video description) which apparently included Irwin Creek for it is mentioned as a fishing site by a pack’s yearling in 2017 (Freund, D.R. et al., 2023) and as in late summer, 2019 it was taken over by a new pack which was then called the Half-Moon Pack (Central location), they also used to reside in the central parts below Lake Kabetogama which possibly indicates that upon developing a strategy of accessing high quality alternative food resources during the perilous periods of early summer and late summer, the pack can thrive even in this middle portion.
Their breeding pair was apparently successful producing ca. 25 pups since 2016 (it is not mentioned how many of these pups survived).
They also survived the territory takeover and after having lost their range, even managed to usurp a territory from the Sheep Ranch Pack which was much bigger.
Unfortunately, the Bowman Bay Pack turned Huron Pack (on the former Sheep Ranch Pack territory) was not long-lived and their new range was taken over by the Paradise Pack (see below).
However, I believe that their relative success is indicative of the problem lying with food availability which they overcame by learning to fish.
Meanwhile, the Paradise Pack utilized blueberries extensively in July – August (i.a., Homkes, A.T. et al., 2020) and the pup diet might have consisted of > 30% blueberries during summer.
VWP even states that some researchers believe berries can make up as much as 80% of a wolf pack’s weekly diet during some years (see video description).
The berry supplements do not appear to have ‘done it’ improving the Paradise Pack’s outlook (if 2 out of 5 pup survival is used as an estimate).
The pack was formed in 2020. As mentioned before, it produced 4 pups in 2020 but all died by July.
In 2021, 2 out of 5 pups survived and the pack even managed to take over the Huron Pack’s territory (although it is possible that this was not an aggressive takeover and that Huron Pack had dissolved due to loss of pups).
Paradise Pack might have better chances now that they could incorporate Huron Pack’s range because their initial territory was very small and crammed among other packs and perhaps could not sustain a family.
According to VWP trap camera videos, blueberries were consumed both in 2020 (video 1) and in 2021 (video 2).
The Voyageurs Wolves are also major beaver hunters with some individuals more expertly than others (i.a., Gable, T.D. & Windels. S.K., 2017).
Beavers have always been a major food source in Voyageurs Ecosystem (15% frequency of occurence in 1987 – 1991 study period).
No numbers have been given for the 2020 – 2022 study period but 42% is the number given in the publication by Gable, T.D. et al., 2020 (possibly estimates of ca. 2017).
This increase could also be indicative of decreased fawn availability and increased reliance on beavers.
Other food items (snowshoe hare, moose, rodents, grasshoppers etc.) have been listed both in the earlier publication and the more recent reports as well as more recent publications on wolf predation behaviour (e.g., Gable, T.D. et al., 2014).
However, these items are consumed opportunistically and rarely compared to deer, beavers, fish and berries.
It has been stated by the VWP team several times that beaver hunting and consumption of alternative food sources during periods of fawn scarcity (late spring / early summer) as well as during period in late summer / autumn when deer fawns have acquired mobility, pups are still developing and needing plenty of energy but harsh weather has not weakened the prey and beavers become less accessible, is normal.
Personally, I disagree and I believe that there is a tendency of decrease in prey abundance (deer fawns) that affects wolves especially in the central packs below Lake Kabetogama and that forced the wolves to employ other food resources but the alternative food source strategy has not been effective enough to prevent high pup mortality rates.
Pup mortality results in a greater pack turnover, social instability, smaller territory sizes creating further pressure on the surviving packs to acquire ever scarcer resources on smaller ranges.
There might be additional factors due to climate change (discussed above) limiting packs’ ability to maintain proper ranges that provide sufficient resources for pup production.
While pack size has decreased both in Kabetogama Peninsula and below Lake Kabetogama, the peninsular packs appear to be affected to a lesser degree wherefore it might be advisable to study recent differences in geographic (especially, hydrological) changes in the two areas as well as prey migration and abundance in these areas.
It might be necessary to decrease deer hunting rates around Voyageurs Ecosystem in order to ensure prey provision for the grey wolves in VNP, proper pup survival and population stability.
It would be important to determine whether there are any obstacles to even deer dispersal and availability of natal sites for fawn production on the wolf range in the central part of wolf packs below Lake Kabetogama.
***
The 1987 – 1991 report also refers to extensive wildfires in the 1920-ies – 1930-ies as well as extensive logging between 1910 and 1930 that shaped the forests in the VNP (also, 25% of Kabetogama Peninsula was logged post 1940).
This could have led to maturing of the forest over time and mature forests tend to accommodate lower ungulate population densities as the canopies close and the understorey vegetation becomes sparser (offering less food and reduced shelter).
Beavers likely maintain structural diversity and openness near the riparian corridors but it is possible that deeper in the forest areas the vegetation communities have undergone succession unfavourable to maintain the former deer densities.
A drop in deer densities could explain the starvation issues in the pup group.
If Kabetogama Pensinsula was logged later, it could presently be found at a later successional stage resulting in yet higher deer densities and subsequent higher wolf survival rates.
Fires, however, have occurred later, as well, affecting, e.g., 0.33% of land from 2002 – 2007 (Mechenich, C. et al., 2015) but the fire frequency appears to have decreased.
***
Currently, the white-tailed deer are the only large wild ungulates available to wolves in VNP.
However, according to Mechenich, C. et al., 2015, historically, also elk and caribou lived on the lands.
This might have once upon a time given a bigger supply of fawns during the calving season although none of the packs the histories of which have been discussed above, experienced this abundance.
Elk and caribou were extirpated from the area by 1920-ies (due to uncontrolled hunting).
It is a possible that while a greater diversity of ungulates might have not necessarily widened the period of fawn availability because in all these species (white-tailed deer, elk, caribou and moose) fawing occurs rather synchronously, it might have contributed to a dispersion of this resource over the ranges , e.g., making it more homogeneously available for all wolf packs.
There can also be differences between fawn rearing behaviour among the species (white-tailed deer and moose can be more solitary during reproductive season than elk and caribou leading to different use of habitats and also to mothers leaving fawns unattended for longer periods of time during early development while the fawn cannot follow the mother).
This might be important in packs with few members where during summer individuals have to hunt solitarily and perhaps, for a while, (in pairs without subordinates) only the breeding male is the dominant provider.
In some earlier publications it has been considered to reintroduce caribou to the VNP.
However, with the climate change, the landscape has turned more temperate than boreal and caribou might not feel at home there any longer.
These reductions in the original large herbivore guild might also bear impact on vegetation structures and succession in Voyageurs Ecosystem.
While it is presumed that the white-tailed deer have increased in abundance since the historic times prior to the extirpation of elk and caribou, they alone (with the scarce numbers of moose) might be unable to promote or to prevent some processes or their particular niche might be a limiting factor.
For example, elk are rather grazers than white-tailed deer are although both species can graze and browse.
Perhaps when elk coexisted with white-tailed deer, the white-tailed deer as the smaller species were outcompeted from more open habitats and thereby encouraged to browse more halting the succession in plant communities.
However, it appears that elk densities have been negligible because they are not even mentioned in most publications apart from that by Mechenich, C. et al., 2015.
Moose, meanwhile, have been far more abundant than currently when they do not constitute significant prey resource for the Voyageurs wolves.
***
I found a partial confirmation of my vegetation succession hypothesis in the white-tailed deer population assessment study by Gable, T. et al., 2017.
I quote, ‘We think the decrease in deer density in VOYA is likely due to changing forest composition and structure. VOYA was last logged in 1971, before the park was established, and little early successional habitat is available in the park, as natural disturbance is largely limited to small-scale events from beaver activity or wind throw.’
This, however, refers to Kabetogama Pensinsula mainly.
***
The study by Gable, T. et al., 2017 also mentions, ‘Kabetogama Peninsula, a 305 km2 roadless areain the center of the park where >90% of the park’s moose population occurs’ – which could explain the comparative success of peninsula wolves that apparently have an additional albeit scarce prey species.
Other important quotes from Gable, T. et al., 2017,
‘Pellet-counts from 1989 and 1991 resulted in deer density estimates of 12.3 ± 1.5 (SE) and 11.6 ± 1.3 (SE) deer/km2, respectively.’
‘In 1996, mid-winter pellet-counts resulted in an estimate of 3.0 deer/km2. However, in 1997 the pellet-count based deer density was 5.2 deer/km2.’
The decline seems rather dramatic although the very low 1996 count could have been a result of a very cold winter with blizzards and substantial windchill that would have affected both deer numbers and pellet count effort.
It is possible that the deer suffered dramatic population reduction during the winters of 1995/96 and 1996/97 that were both extremely harsh.
I do not currently recall any publications discussing how much time is needed for an ungulate population to recover after devastating losses due to extreme weather but I suppose that > 25 years should have brought the numbers up if there were no additional factors at play to keep them low.
But see, ‘We cannot fully explain why pellet-based density estimates were so high during 1989–1991, but it may have been due to methodological errors, such as incorrectly classifying old pellets as new ones, or random sampling errors, such as if several sample units fell within existing deer yards.’
The drop in density might have not been as dramatic if the estimates of 1989-1991 were imprecise.
Adjacent to the park, officials reported densities that were ca. 50% lower (which would fall closer to 1997 count) but outside of the park, deer were also hunted.
This is another factor that should be considered when comparing the Kabetogama Peninsula and below-Lake Kabetogama populations because it appears that on the peninsula the deer are not as affected by the hunting pressure that impacts the deer that can wander outside of the protected area (keeping in mind that deer are philopatric with relatively small home ranges and restricted movement outside of winter migrations to deer yards).
Also, it is noted that white-tailed deer density might have been inflated in the 1980-ies to begin with not reflecting abundances that would be characteristic to a more stabilized situation, ‘It is possible that deer abundance in VOYA experienced a significant increase for a decade or so after the park was established in 1975, resulting from a combination of protection from legal hunting and a large pulse of early successional habitat created in the 1950–1960s.’
‘In 2009, deer density (based on pellet-counts) was 4.1 ± 0.5 (SE) deer/km2, but declined to 1.71 ± 0.4 (SE) deer/km2 by 2016.’
‘There was no difference (P>0.05) in deer density estimates inside and outside of VOYA during 2009, 2010, 2015, and 2016. Similarly, there was no difference (P>0.05) in deer density estimates on the Kabetogama Peninsula and the area south of Kabetogama/Namakan Lake (i.e., southern portion of VOYA, and state/federal land south of VOYA).’
(However, the estimates were made on 1km2 sampling plots (for pellet counts) which might be a scope that is too large to detect variation in localized distribution among wolf pack home ranges. Also, the effort was not as concentrated below Lake Kabetogama (outside of Voyageurs NP) than on Kabetogama Peninsula.)
‘Of particular note, two severe winters ending in 2013 and 2014 preceded the low deer density estimate of 2015, though we would have expected the 2016 deer density estimate to have been higher than 2015 based on the two mild winters of 2015 and 2016.’
***
I had thought that the Voyageurs Ecosystem was a nature preserve also outside of the VNP but it appears that logging occurs in the Ecosystem.
‘Either way, this new pack, who we named the Clearcut pack after the extensive logging operations within the territory, was now clearly occupying the former Bluebird Lake territory.’ (Voyageurs Wolf Project Report, 2022 – 2023).
In the same report, map is given (pp. 7) indicating borders of the Voyageurs National Park and the Greater Voyageurs Ecosystem.
The specific logging operation occurred outside of the VNP and slightly to the SE.
The packs tend to be larger there and their territories bigger in size.
Meanwhile, the packs that have struggled, seem to reside close to VNP south border or slightly overlapping it.
I wonder if this indicates at the effect that recent logging operations might have on deer distribution patterns.
Perhaps deer are more abundant near clearings as it is often observed in different types of ecosystems.
On a population level, deer might also concentrate toward the areas where forage is more abundant.
Perhaps the logging south of VNP draws some deer out to the clearings.
White-tailed deer can be migratory but even if they do not emigrate from the VNP southern range (below the Kabetogama Peninsula), the distribution on a population level might have become skewed toward the more ‘optimal’ areas.
And if they are migratory, perhaps this could explain why pup survival in the southern range of VNP is poor in late summer/autumn/early winter as the fawns can follow their mothers and they migrate to more abundant ranges.
These packs below the Kabetogama Peninsula have small territories and they might not be able to follow the deer distribution changes or to compensate for the distribution skew on the population level.
Of course, I am not informed on where logging actually occurs. Perhaps it also occurs on the territories of those packs that reside on the boundary between the VNP and the GVE.
‘
Notes by Ieva 