Productivity of semi-domesticated reindeer in Finland

In spite of a twofold increase in the density of reindeer in Finland from 1974 to 1987, meat production per reindeer increased during this period. This was possible due to calf harvesting and supplemental feeding. Results from multiple regression models indicated that calf harvesting influenced both per unit area and per capita production more than supplemental feeding. Correlation between meat production and animal density decreased with increased supplemental feeding. Traditionally, southern and central herds of reindeer fed mainly on arboreal lichens in late winter; however, due to large-scale logging, woodlands rich of arboreal lichens had been greatly reduced. Economic carrying capacity of the winter range apparently has been exceeded in the south; a heavy crash in the number of reindeer is likely if supplemental feeding ceases. In northern herds, intensive calf harvesting enabled satisfactory yield without supplemental feeding. In northern herds, yield increased mainly per unit area (i.e. by increases in herd size); in the south yield per reindeer increased. _ t 1 c ~ r ~ Rangifer, l l (2): 53—63 K e y words : reindeer, meat product ion, F i n l a n d , calf harvest, supplemental feeding Kojola, I., Helle, T. & Aikio, P. 1991. Lithantuotantoon vaikuttavat tekijat Suomen poronhoidossa. Tiivistelmd: Huolimatta Suomessa vuosien 1974 ja 1987 valilla tapahtuneesta porotiheyden kaksinkertaistumisesta, l i hantuotto poroa kohti kasvoi jakson aikana. Tama johtui oletettavasti vasateurastuksesta ja lisaruokinnasta. Monimuuttujaregressiomallien tulosten perusteella vasateurastuksella nayttaisi olevan lisaruokintaa suurempi vaikutus seka poroa etta pinta-alaa kohti laskettuun tuottoon. Ruokinnan tehostuessa pienentyi lihantuoton ja porotiheyden valinen riippuvuus. Etelaosan ja keskiosan porot syovat perinteisesti puussa kasvavia jakalia kevattalvella. Hakkuista johtuen luppometsien osuus on suuresti vahentynyt. Talvilaidunten ekonomien kantokyky on ilmeisesti ylitetty etelaja keskiosassa; syva romahdus poromaarissa on todennakoista, jos ruokinta lopetettaisiin. Pohjoisosassa voima-perainen vasateurastus mahdollistaa tyydyttavan tuoton ilman ruokintaa. Pohjoisessa tuotto kasvoi pinta-alayksikkoa kohden (poromaarat kasvoivat), etelassa kasvoi poroa kohti laskettu tuotto. Rangifer l l (2)* 53 63 Introduction most preferred winter food, terricolous lichens A t the equi l ibr ium of economic carrying capacity, (Cladina spp.), is far below the biomass that provithe opt imal animal density permits m a x i m u m prodes m a x i m u m product ion of l ichen (Helle et al.s duction (Caughley, 1976). H i g h animal densities 1990). Reindeer can part ly compensate for low biand consequent range deterioration are c o m m o n l y omass of terricolous lichens by using arboreal associated w i t h intensive reindeer management, and lichens or vascular plants (Laws, 1981; Leadermay result i n increased juvenile mortal i ty and low W i l l i a m s , 1988), but shortages of lichens can increacarcass weights (Skogland, 1983; 1985). In the reinse l ike l ihood of populat ion crashes dur ing severe deer management area of F in land , biomass of the winters. Occasional mass-deaths and consequent Rangifer, l l (2), 1991 53 large-scale fluctuations in reindeer numbers were c o m m o n i n F i n n i s h reindeer management during the 1960's and in the early 1970's. D u r i n g 1974-1987, no substantial crash occurred, and the density of reindeer doubled without an increase in mortal i ty or decrease in reproductive rate. Two plausible reasons exist for such a development (Helle & Santti, 1982; Hel le & Kojola , 1991). First , maleskewed calf harvest has reduced the number of calves and males, both of w h i c h are more susceptible than females to starvation in winter (Kle in , 1968; LeaderW i l l i a m s , 1980; Skogland, 1985). Second, supplemental food was provided in southern and central herds, where number of reindeer per range suitable for terricolous lichens was higher and biomass of lichens consequently lower than i n northern herds (Helle et al, 1990). M a x i m u m meat product ion is the main objective of F i n n i s h reindeer management. Here we evaluate meat product ion per unit area and per reindeer at different levels of animal density, supplemental feeding, and variable calf harvesting. Study area and methods Study area. There are 56 herding associations (herds) i n the reindeer management area of F i n l a n d (F ig : 1). Areas of herding associations varied f rom 558 to 5580 k m 2 . Traditionally, reindeer i n the northern part of the management area seek for food f rom below the snow during the whole winter, whi le in the south shift to arboreal lichens during the course of winter. We divided herds into 3 regional groups by the length of the time during w h i c h reindeer obtain their forage f rom beneath the snow (Fig. 1). In the northern region this period is 7 months, i n the central region appr. 5 months and i n the southern region ca. 3 months (Helle & Saastamoinen, 1979; H e l l e & Tarvainen, 1984). Densities of reindeer doubled dur ing 1974-1987; the increase was most pronounced i n northern herds (Helle & Kojo la , 1991). The government of F i n l a n d sets the m a x i m u m number of reindeer for each herding association every 10 years; i n northern and central herds numbers were comm o n l y exceeded i n the 1980's. The amount of hay provided per reindeer increased f rom nor th to south. Study period. Here we evaluated the period f rom 1960 to 1987. Cal f harvesting began i n some central herds i n the m i d 1960's. Af ter a major populat ion crash i n 1973 calf harvesting was intensified also i n the nor th (Fig. Fig. 1. Reindeer management area of Finland, divided into subareas used in this study. The period of digging for food from beneath the snow: northern herds 7 months, central herds 5 months, and southern herds 3 months. 2). Supplemental winter feeding also began after this crash. The amount of supplemental food provided per reindeer was greater during the 1980's than 1970's (Fig. 2; Hel le & Saastamoinen, 1979; N i e m i ¬ nen & Aut to , 1989). W h e n comparing herd means, we divided our study per iod further: (1) 1960-1966 (very few calves harvested, no supplemental winter feeding) (2) 1967-1973 (some male calves harvested, no supplemental feeding, decreasing populations), (3) 1974-1980 (some calf harvest and supplemental feeding, increasing populations) (4) 1981-1987 (increased calf harvest and supplemental feeding, increasing populations). We calculated intraherd regressions f rom 1960 to 1973 and f rom 1974 to 1987. Data collection. We extracted numbers and harvest rates of reindeer f rom the archives kept by the Associat ion of Reindeer Herders, and carcass weights of reindeer f rom the sales books maintained by meat buyers. Data com54 Rangifer, 11 (2), 1991


Introduction
most preferred winter food, terricolous lichens At the equilibrium of economic carrying capacity, (Cladina spp.), is far below the biomass that provithe optimal animal density permits maximum pro-des maximum production of lichen (Helle et al.s duction (Caughley, 1976).High animal densities 1990).Reindeer can partly compensate for low biand consequent range deterioration are commonly omass of terricolous lichens by using arboreal associated with intensive reindeer management, and lichens or vascular plants (Laws, 1981; Leadermay result in increased juvenile mortality and low Williams, 1988), but shortages of lichens can increacarcass weights (Skogland, 1983;1985).In the rein-se likelihood of population crashes during severe deer management area of Finland, biomass of the winters.Occasional mass-deaths and consequent large-scale fluctuations in reindeer numbers were common in Finnish reindeer management during the 1960's and in the early 1970's.During 1974During -1987, no substantial crash occurred, and the density of reindeer doubled without an increase in mortality or decrease in reproductive rate.Two plausible reasons exist for such a development (Helle & Santti, 1982;Helle & Kojola, 1991).First, maleskewed calf harvest has reduced the number of calves and males, both of which are more susceptible than females to starvation in winter (Klein, 1968;Leader-Williams, 1980;Skogland, 1985).Second, supplemental food was provided in southern and central herds, where number of reindeer per range suitable for terricolous lichens was higher and biomass of lichens consequently lower than in northern herds (Helle et al, 1990).
Maximum meat production is the main objective of Finnish reindeer management.Here we evaluate meat production per unit area and per reindeer at different levels of animal density, supplemental feeding, and variable calf harvesting.

Study area.
There are 56 herding associations (herds) in the reindeer management area of Finland (Fig: 1).Areas of herding associations varied from 558 to 5580 km 2 .Traditionally, reindeer in the northern part of the management area seek for food from below the snow during the whole winter, while in the south shift to arboreal lichens during the course of winter.We divided herds into 3 regional groups by the length of the time during which reindeer obtain their forage from beneath the snow (Fig. 1).In the northern region this period is 7 months, in the central region appr.5 months and in the southern region ca. 3 months (Helle & Saastamoinen, 1979;Helle & Tarvainen, 1984).Densities of reindeer doubled during 1974-1987; the increase was most pronounced in northern herds (Helle & Kojola, 1991).The government of Finland sets the maximum number of reindeer for each herding association every 10 years; in northern and central herds numbers were commonly exceeded in the 1980's.The amount of hay provided per reindeer increased from north to south.

Data collection.
We extracted numbers and harvest rates of reindeer from the archives kept by the Association of Reindeer Herders, and carcass weights of reindeer from the sales books maintained by meat buyers.Data com-  1966 1967-1973 1974-1980 1981-1987 Fig. 2. The amount of supplemental hay provided per reindeer and the proportion of calves harvested (% of total number of calves) (means ± SD) in different parts of the reindeer management area of Finland.
prises of 1568 cases.The amounts of supplemental hay provided per reindeer were calculated from data published by Helle & Saastamoinen (1979) and Nieminen & Autto (1989).Round-ups and harvesting took place in early winter.Alaruikka (1964) estimated that in the early 1960's the proportion of un- i_ O Ü Fig. 3, During the late 1970's, the proportion of unmarked reindeer was 1-2% of all reindeer on the average (Autto, 1980;Niittyvuopio, 1981).We calculated per capita meat production by dividing the total meat production by the number of reindeer left after the round-up of the previous year.Round-ups and harvesting took place in early winter.When conducting analysis on the profitability of feeding, we used a producer price of 30 FIM kg -1 meat (the actual price in the mid 1980's) and 1 FIM kg -1 as a price of hay.
Statistical analysis.
Regional differences in meat production were tested by using oneway analysis of variance.The dependence of per capita and per unit area meat production on reindeer density, supplemental feeding and the rate of calf harvesting was tested by using simple and multiple regression analysis.To test whether

Production per unit area
Inrease in meat production per unit area from the period 1974-1980 to 1981-1987 was most pronounced in northern herds, and reached a value higher than elsewhere (Table 1).Interherd correlation between production and reindeer density a^rjeared to fluctuate less in southern than in central or northern herds (Fig. 3).In 1974-1980 calf harvesting influenced production per unit area only in the southern region, while in 1981-1987 production depended on calf harvesting both in the southern and northern region (Table 2).For the entire range, calf harvesting was influental during both of these periods (Table 2).Supplemental feeding did not affect significantly production in any multiple regression models fitting results for meat production per unit area (Table In 1960-1973, inter-year variation of meat production within herds was correlated positively (P < 0.05) with density in 34.5% of all herds (n = 56).In 1974In -1987  was significant in northern (P < 0.001, n = 11) and central (P < 0.002, n = 21) herds but not in southern herds (P = 0.386, n = 24).During 1974-1987, intraherd correlation between meat production and animal density was negatively correlated with the amount of hay given per reindeer during the winter (Fig. 4).This correlation also was correlated significantly with the mean rate of calf harvesting (r = -0.386,n = 56, P < 0.001).

Discussion
Productivity and range conditions Data on production are available for comparison only from hunted populations of Norwegian wild mountain reideer (Gaare & Skogland, 1980;Skog¬ land, 1986), which live mostly in open habitats and are of the same genetic origin as the wild ancestors of Finnish semi-domesticated reindeer (Siivonen, 1975;Nieminen & Helle, 1980).In wild reindeer, the maximum sustained yield of 14 kg meat km -2 is obtained at a population density of 1.75 reindeer km -2 (Skogland, 1986).These estimates suggest that without earlier over-grazing, the area associated with subtle lichen forage, comprises 12% of the total herd land area.
In southern and central herds in our study area, the proportion of area suitable for lichen was similar to that in Norway.In the 1960's and 1970's, before the initiation of intense supplemental feeding, meat production per unit area was substantially lower in the central and southern ranges than in Norway.Despite a lower animal density (Helle etal., 1990), production exceeded the Norwegian level in the 1980's.In the north, where the Norwegian level was clearly exceeded during the 1980's, the winter range: total range ratio was considerably higher than that in Norway (Mattila, 1981).In wildlife systems herbivore abundance is often controlled by the amount of high quality food in a critical season (Sinclair, 1975;Skogland, 1985).The basic difference between the Norwegian model and Finnish practice is the condition and role of lichen ranges.In Norway, a gross density of 1.75 reindeer km -2 allows maximum lichen production, providing reindeer with terricolous lichens to feed on the whole winter season.In southern and central herds in Finland, scarcity of terricolous lichens force reindeer to forage on arboreal lichens in mid and late winter.Lichen ranges were heavily overgrazed already in the beginning of this century (Anon., 1914).In northern herds overgrazing was reported in the 1950's and 1960's (Andreev, 1971).By the 1970's overgrazing in the north had reduced the biomass far below necessary for maximal rate lichen prodution (Helle etal, 1990).Substitute foods for lichens in mountain areas consist mainly of sedges, grasses and dwarf-shrubs (Helle, 1984).
Negative correlations between density and per capita meat production within herds in the periods 1960-1966 and 1967-1973 were obvious indications of food limitation.Each herd experienced several crashes in the 1960's and 1970's (Helle & Kojola, 1991).The crash was followed by a period of increase in the herd size until the next crash.As a consequence, animal density peaked prior to a year of high mortality and low productivity.This may explain why meat production per unit area did not correlate with density in 1960-1973.A proximate reason for the crashes was adverse snow conditions (Helle & Sàntti, 1982), but the ultimate reason was densitydependent food-limitation.
Positive correlation between per capita production and reindeer density in 1974-1980 and 1981-1987 was no doupt partly due to exceedings of the maximum permitted number of reindeer: when herd size outnumbered the legal limit, a higher proportion of reindeer was culled.In 1960In -1966, when , when the relationship between per capita production and density were negative in each region, the size of the herds was usually below the number maximally permitted.
The improvement of the condition of lichen ranges by means of reducing reindeer density has not been considered as a realistic option in southern and central herds.An estimate of the biomass is 10 tons km -2 (Mattila, 1988), whilst in its most productive condition it would be 9o tons km -2 (Karenlampi, 1973).With an annual growth rate of 11%, achieving maximum lichen production would take about 30 years.Without perfect exclusion of grazing, this process would be longer.The linear type functional response of reindeer to lichen supply (White & Trudell, 1980;Skogland, 1980) suggests that when lichen vegetation begins to recover due to lowered density, reindeer shift to forage increasingly on lichens and thereby slow the growth of the lichen.Terricolous lichens are of greatest importance in northern herds because arboreal lichens are absent there.Because of higher initial biomass, 25 tons km -2 (Mattila, 1981) in the north, recovery would be faster thanin southern and central herds.The heavy population crash in 1973 evoked awareness about range conditions and some herding associations reduced the highest permitted number of reindeer for the 1980's.However, the permitted herd sizes were exceeded by the middle of the decade, and the excesses were greatest in the north.
Supplemental feeding and calf harvesting Supplemental feeding and calf harvesting were a more or less conscious response to range deterioration.In southern and central herds this was accelerated by a reduction of the arboreal lichens as a result of large-scale logging of old forests (Helle & Saastamoinen, 1979).In contrast to economically managed herds of reindeer (see Gaare & Skogland, 1980;Skogland, 1986), the main objectives of the herder in traditional pastoralism are not maximal yield and income but rather a subsistence supply and a large herd (Noy-Meir, 1981).In reindeer management, a large herd provides its owner with power and influence within the herding association (Ingold, 1980).In the 1960's and 1970's, reindeer owners shifted from a largely subsistence based way of life to a monetary economy that stresses stable income (Pelto, 1973;Ingold, 1980).A predictable annual yield is required in modern reindeer husbandry but cannot be achieved without careful management when winter ranges are poor of most preferred forage.Although the supplementary feeding and changes in the structure of the winter herd were known, doubling of the number of reindeer and production of meat was unpredictable.Our results indicate that in southern herds increased production was more clearly related to increased yield per reindeer than in central and northern herds, where the increase was more clearly related to increased reindeer density.
Effects of supplemental feeding and calf harvesting were difficult to distinguish decisively from each other.They both were initiated in southern herds and then intensified and spread towards the north at the same rate.Results from the multiple regression analysis used to compare these effects indi-cate that calf harvesting has a greater influence than feeding on production.These results are not, however, very conclusive because our model did not take into account that without supplemental feeding animal density should be substantially lower in southern and central herds.Because supplemental feeding at the same time compensated for the loss of natural food resources, it is impossible to isolate its effects.In addition, it is fair to assume a dependence between these independent variables: supplemental feeding enables high calf/female ratio (Helle & Kojola, 1991), which in turn is necessary for extensive calf harvest.In the southern region most reindeer owners are farmers and therefore able to obtain high per capita production by supplemental feeding at reasonable costs.In north herds, some feeding is necessary when snow conditions are exceptionally difficult.As stated by Noy-Meir (1981), intensive supplemental feeding has also considerable ecological consequences: «Where supplementary feeding becomes economically feasible this may induce further increases in the herd and lead to maintenance of a highly overabundant herbivore population on a severely overgrazed range.»Harvesting of male calves was introduced into Finland from Russian experimental herds, and its original objective was to make space for females in winter herd (Alaruikka, 1964).With regards meat production, the resultant sex ratio (commonly 1 male to 10 females) provides the most feasible option (Arobio et al, 1980).Since the mid 1970's in the south and the 1980's in the north, harvesting of female calves became necessary.Otherwise the killing of females in their best reproductive age would be the only way to keep the number of reindeer within legal limits (Lenvik, 1989).
We suggest that intensified calf harvesting is the most important reason for the recent increase in production in northern herds.Calf harvesting is a tool enabling maximal production of meat from a herd of reindeer (Arobio et al, 1980).This results from the rapid growth of calves and their susceptibility to starvation in winter (Skogland, 1985).In an earlier paper we put forward a hypothesis that there exists another mechanism by which calf harvesting increases productivity (Kojola& Helle, 1991a).This was based on the finding that calf harvesting had a distinct positive effect on the subsequent year's calf/female ratio in northern herds, where reindeer dig for food from beneath the snow the whole winter.Calves commonly share feeding craters with their mother and thus the food made available by the mother (Kojola, 1989).The nutritional status of breeding females is therefore likely to improve if their calves are killed before winter (Kojola & Helle, 1991a).This furthermore may bring about better survival of subsequent year's calf: close relationship between maternal condition and offspring survival is well documented in reindeer (Rognmo etal, 1984;Lenvik 1989).Despite low energy content and poor digestibility of sedges, grasses and dwarf-shrubs, females may overwinter in satisfactory condition when most calves are killed before winter.In northern herds, calf harvesting frees reindeer from density-dependent mortality and recruitment rate (Kojola & Helle, 1991b;1991c).Tjime trends in carcass weights indicate a slight decrease in northern herds (Helle etal, 1991).
In conclusion, there was no evidence of exceeding of the economic carrying capacity in the northern region although animal density in some herds reached figures higher than so far reported for any continental population (Helle & Kojola, 1991).The concept of carrying capacity is based on the ratio between natality and mortality (Caughley, 1976).Like most vertebrates, the mortality rate of reindeer depends on age and sex (Klein, 1968;Leader-Williams, 1980;Skogland, 1980;Reimers, 1983).A lower initial mortality and higher reproductive rate is achieved, and consequently, higher carrying capacity of the range was created in the northern region, by means of altering sex and age ratios in the winter herd.

Fig. 1 .
Fig. 1.Reindeer management area of Finland, divided into subareas used in this study.The period of digging for food from beneath the snow: northern herds 7 months, central herds 5 months, and southern herds 3 months.

Fig. 4 .Fig. 5 .
Fig. 4. Relationship between correlation of meat production per unit area with reindeer density and the intensity of supplementary feeding in 1974-1987 in Finland.Each point represents one herd.Triangles indicate northern, circles central and quadrats southern herds (Y = 0.823 -0.003 x X, r = -0.570,n = 56, P < 0.001).

Table 1 .
Meat production per unit area in different parts of the reindeer management area of Finland.

Table 2 .
Multiple regression models fitting results for meat production per unit area (kg/km 2 ) and the analysis of variance for the full regression.

Table 3 .
Meat production per reindeer in different parts of the reindeer management area of Finland.

Table 4 .
Multiple regression models fitting results for meat production per reindeer (kg) and the analysis of variance for the full regression.Density: Winter density (reindeer/km 2 land area); Calf harvesting: the proportion of calves harvested (% of total number of calves); Supplemental feeding: the amount of supplemental hay (kg/reindeer).