THE SUPREME COURT OF APPEAL OF SOUTH AFRICA
JUDGMENT
Not Reportable
Case no: 316/2022
In the matter between:
SUMMERMANIA ELEVEN (PTY) LTD APPELLANT
and
WILLIAM HENRY HATTINGH N O RESPONDENT
Neutral citation: Summermania Eleven (Pty) Ltd v Hattingh N O (316/2022)
[2024] ZASCA 42 (5 April 2024)
Coram: DAMBUZA ADP, SCHIPPERS, MBATHA, MOTHLE and
GOOSEN JJA
Heard: 03 May 2023
Delivered: This judgment was handed down electronically by circulation
to the parties’ representatives by email, published on the
Supreme Court of Appeal website, and released to SAFLII.
The date and time for hand -down is deemed to be 11h00 on
5 April 2024.
Summary: Contract law – breach of a warranty – onus on the plaintiff to prove
breach.
Civil procedure – in a trial, evidence is given orally – court may grant an order
that evidence be given by way of affidavit – the probative value of evidence given
2
by way of affidavit should be evaluated against the evidence as a whole –
evidence insufficient to discharge onus on the appellant.
3
ORDER
On appeal from: Eastern Cape Division of the High Court, Makhanda: Lowe J,
with Roberson and Bloem JJ concurring) sitting as a court of appeal:
The appeal is dismissed with costs.
JUDGMENT
Dambuza ADP (Schippers, Mbatha, Mothle, and Goosen JJA concurring)
[1] At the heart of this appeal is the correct approach to evaluation of evidence
in civil trials. The Eastern Cape Division of the High Court, Makhanda (trial
court, Mfenyana AJ) admitted into evidence an affidavit deposed to by the
appellant’s expert witness and found, on the strength of that affidavit, that the
respondent had breached a term of a sale agreement concluded between the
parties. A full court of that Division (Lowe J with Roberson and Bloem JJ
concurring) reversed the decision of the trial court, and found that the trial court
did not properly evaluate the expert evidence. This appeal, against the judgment
of the full court, is with the special leave of this Court.
[2] The dispute between the parties emanates from a sale agreement concluded
between them o n 3 F ebruary 2015. In terms of that agreement the appellant,
Summermania Eleven (Pty) Ltd (Summermania), bought from the respondent,
the Billy Hattingh Trust ( Trust) a game farm located in Graaff-Reinet, in the
Eastern Cape Province . In terms o f the written agreement of sale the farm was
bought as a going concern. T he purchase price included game on the farm. The
composition and estimated numbers of the various game species w ere set out in
an addendum to the sale agreement. The addendum was prepared in
4
December 2014. In terms of Clause 5 of the agreement, the Trust guaranteed that
there would be no material change to the composition and numbers of game as at
December 2014 . There was a further warranty in terms of which the Trust
undertook to maintain ‘the property’ in the same conditio n it was in December
2014, until registration of transfer to Summermania.
[3] Transfer of the property in the name of Summermania was registered on
10 July 2015 , much later than the parties had anticipated . Mr Enrico Nielsen ,
Summermania’s sole director, took occupation of the farm on 15 July 2015. A
year later, o n 11 July 2016 , Summermania instituted proceedings against the
Trust claiming damages for breach of contract in a number of respects. Relevant
to this appeal is t he allegation that the Trust breached its undertaking to ensure
that there would be no material change in the composition and numbers of game.
Regarding this claim the Trust merely denied that it had breached the warranty
given in the Deed of Sale.
[4] Summermania alleged that after taking occupation Mr Nielsen discovered
that there was significantly less game on the farm than had been set out in the
addendum to the sale agreement. In the summons Summermania pleaded that the
numbers of the Nyalas had decreased by seven, the Bushbuck had decreased by
eight, the Waterbuck, by 16, the Zebras, by 14, the Mountain Reedbuck, by 20,
the Vaal Reedbuck, by seven, and the Eland by seven. Most concerning to
Summermania was the reduction in the number of Kudus on the farm by 150.
Although the amount claimed as damages included the value of the other missing
animals, the dispute and the contested expert evidence was centred around the
missing 150 K udus. Summermania contended that bec ause of their size the
missing Kudus could not have gone unnoticed by the Trust’s representatives.
5
[5] On the undisputed facts, the alleged reduction in the given game numbers
would have happened over a period of about 18 months from the conclusion of
the sale. It took another four months from the date of occupation for Mr Van
Niekerk to conduct the game count in November 2015. The November game
count report showed that some of the numbers of other game species had
increased.
[6] The report prepared by Mr Benjamin Van Niekerk, an expert game counter,
formed the basis of Summermania’s claim. However, Mr Van Niekerk could not
give evidence at the trial because of ill health. His evidence was submitted to
court in the form of an affidavit deposed to on 9 January 2020, five years after he
had done the game count.
[7] The trial court admitted the affidavit in evidence, and on the strength of
that evidence, found that the Trust was in breach of the agreement ‘in failing to
comply with its obligations in terms of clause 5 of the agreement’. In reversing
that decision the full court found that the trial court had not properly evaluated
Mr Van Niekerk’s evidence. The full court reasoned that Mr Van Niekerk’s
evidence should have been approached with caution, and, should not have been
‘accepted’ given the nature of his illness and the fact that it could not be tested in
cross-examination. The court reasoned that there was no evidence of any factor,
such as drought or disease that would have affected the composition and numbers
of game to the extent deposed to by Mr Van Niekerk. Consequently, the inherent
probabilities we re that the composition and numbers of game would have
remained the same between December 2014 and date of transfer of the farm to
Summermania.
[8] In this appeal, Summermania contends that the full court erred in several
respects. The court erred in placing on Summermania the burden of proving the
6
probable cause for the reduction in the game numbers, because that was not within
its knowledge. It was submitted that because Mr Nielsen excluded all major
factors such as theft, drought, disease, hunting, and similar dangers that could
have caused reduction of the game numbers, Summermania had discharged the
onus to prove that there was a breach of the warrant y, and that the game must
have gone ‘missing’ prior to Mr Nielsen’s occupation of the farm in July 2015.
There was also a veiled suggestion th at the game numbers recorded in the
addendum may have been overstated.
[9] Clause 5 of the Deed of Sale states:
‘The seller warrants that there will be no material change in the game numbers or game
composition on the property as from the date of inspection being December 2014. The seller
will provide on registration to the purchaser all permits related to the property including
Certificate of Adequate Enclosure.’
There was no dispute between the parties as to the meaning of this clause. The
suggestion that there could have been an overstatement in the addendum must be
dismissed. It must be accepted that when the agreement was concluded both
parties were satisfied with the accuracy of the composition and numbers of game
as recorded in the addendum. In any event, Summermania never pleaded that the
Trust misrepresented the game numbers in the addendum.
[10] As the giver of the warranty the Trust guarant eed a specific state of
affairs – the maintenance of the game numbers as they were in December 2014.
Summermania, having alleged that the guaranteed state of affairs had not been
maintained bore the burden of demonstrating, through credible and reliable
evidence, that this was indeed so. Summermania purported to discharge the onus
on it through the evidence of Mr Nielsen, Mr Nathan Regal and Mr Van Niekerk.
7
[11] Mr Nielsen’s evidence did not take the matter anywhere . I n essence he
related how, soon after taking occupation of the farm during July 2015 he noticed
that the numbers of the game were less that those set out in the addendum. Hence,
he approached Mr Van Niekerk to conduct the game count.
[12] Mr Regal’s evidence, on the value of the missing game, was similarly of
little assistance. He described his occupation as a professional hunter, a game
capturer, a taxidermist, and a trader in game, with over 2 0 years’ experience in
the industry (since 1992) . His evidence was based on his expertise in game
valuation. However, during cross-examination he responded to questions relating
to efficiency of aerial game counts.
[13] Mr Van Niekerk ’s evidence on the aerial count was essential for
Summermania to demonstrate the redu ction in the number of game . He gave
evidence as an expert in game counting, game capturing and trading in game. The
difficulty is that his evidence was required at a very unfortunate time . He was
seriously ill at the time of the trial, suffering from a neuro -degenerative disease
that caused him discomfort when seated for long. He was nearing the end of his
life and his condition caused him to be emotionally fragile.
[14] When called to testify, Mr Van Niekerk’s evi dence lasted a few minutes.
Because of his condition, he began to cry uncontrollably and was unable to
recover. The trial was postponed for some months, until Summermania brought
an application to tender Mr Van Niekerk’s evidence by way of an affidavit. Dr
Marcell Britz who filed a medical report in support of the application, explained
that although Mr Van Niekerk could reflect on past events , testifying in open
court would cause him anxiety, such that he would break down. Despite the
Trust’s opposition to the application , the trial court allowed Mr Van Niekerk’s
evidence to be presented by way of an affidavit.
8
[15] Uniform Court Rule 38(2), in terms of which Mr Van Niekerk’s evidence
was tendered, provides that:
‘The witnesses at the trial of any action shall be orally examined, but a court may at any time,
for sufficient reason, order that all or any of the evidence to be adduced at any trial be given on
affidavit or that the affidavit of any witness be read at the hearing, on such terms and conditions
as to it may seem meet: Provided that where it appears to the court that any other party
reasonably requires the attendance of a witness for cross-examination, and such witness can be
produced, the evidence of such witness shall not be given on affidavit.’
[16] In trial proceedings parties discharge the onus on them by giving oral
evidence. The words ‘shall be orally examined’ in rule 38(2) affirm this as the
standard procedure. Reliability and credibility of the evidence given is then
assessed and analysed through cross-examination. The trial court, however, has a
discretion under the rule, to receive evidence given by way of affidavit. 1 In this
case, the court, in the exercise of its discretion, admitted the evidence of Mr Van
Niekerk. It accepted that Mr Van Niekerk was unable to testify in open court
because of his medical condition.
[17] It was submitted on behalf of the Trust that Mr Van Niekerk’s evidence
should have been rejected by the court, because of discrepancies in that evidence,
and because Mr Van Niekerk could not be cross -examined. The argument on
behalf of the Trust was not that the court exercise d its discretion improperly or
injudiciously. It was also not the Trust’s case that Dr Britz’ evidence and opinion
on Mr Van Niekerk’s medical condition was incorrect or unacceptable. It had to
be accepted, therefore, that Mr Van Niekerk could not give oral evidence and be
cross-examined. His affidavit was therefore properly admitted. The admission of
the evidence nevertheless required t he court to consider whether it was
the evidence nevertheless required t he court to consider whether it was
1 18 Lawsa 3 ed para 200.
9
sufficiently reliable and credible and whether the evidence given thereby was
sufficient, when considered against the rest of the evidence led at the trial , to
discharge the onus on Summermania to prove the allegation of significant
reduction in game numbers.
[18] Regarding his expertise , Mr Van Niekerk described his experience in
animal behaviour, and asserted his familiarity with the topography of the area in
which the farm is located. He acquired his experience in game counting over a
period of 25 years of farming in game and 15 years ‘in the field of game counting
and game capturing’. He explained that because of his experience he knew
exactly where, in the bush, the game would hide on being unnerved by the sound
of the helicopter during the count, and which species would stay calm and remain
in the herd. He had conducted more than 50 ‘commercial serial game counts’ and
estimated his accuracy rate at 80%.
[19] Mr Van Niekerk’s evidence related to the method he employed and his
observations when conducting the count . Regarding the method used, he
explained that he conducted the aerial game count on a clear and sunny day,
sitting in a helicopter which was flown over the farm along north-south and east-
west grid lines which were 300 meters apart. Only he and the helicopter pilot
were in the helicopter. They flew over the farm at an altitude of 300 to 600 meters,
covering the valleys several times, with pauses in between the grid direction
changes, to make allowance for animals to come out of hiding.
[20] As to what he was able to observe, Mr Van Niekerk highlighted the size of
the Kudus, stressing their weight of up to 300kg, a shoulder height of 1.4 meters,
and horns of up to 1.8 meters. This made them easier to spot from the air,
especially because in Novembe r 2015 the Eastern Cape had been experiencing
10
significant drought and the farm was not as dense ly vegetated as usual. He was
certain that he could not have missed more than 30 Kudus during the count.
[21] Mr Van Niekerk acknowledged that during the count some Kudus did hide
in the bush. He counted 50 Kudus (compared with 200 as stated in the addendum)
and he made an allowance for an additional 30, which he thought was generous,
as he could not have missed that number of Kudus given the circumstances at the
time of the count.
[22] Considered on its own , Mr Van Niekerk’s evidence appears simple and
credible, leaving an impression that when he conducted the count, he used a
method with acceptable safeguards, and the result should be reliable. However,
its reliability could not be tested due to the manner in which the evidence was
given. The trial court had to consider that the Trust was denying all material
aspects of Summermania’s case: namely, the fact that Mr Van Niekerk conducted
a game count, the integrity of the methodology of the count , if it did occur, and
the result of the count. A number of concerns arise when considered together with
the rest of the evidence led at the trial.
[23] In his evidence Mr William Henry Hatting h, the Trust’s sole trustee and
witness at the trial , maintained that h e had counted 200 K udus on the farm
throughout 2015. He insisted that, although he and his team had continued with
hunting operation s on the farm subsequent to the conclusion of the sale
agreement, the game numbers did not change. Not much criticism was raised in
respect of Mr Hattingh’s evidence.
[24] Contrary to Mr Van Niekerk’s description of the terrain on the farm , Mr
Hattingh described the topography of the farm as rugged and consisting
predominantly of mountainous areas with deep valleys, and only a small plateau
11
on one side, such that it was ‘too dangerous’ to do an aerial game count on the
farm. This aspect was not mentioned in Mr Van Niekerk’s evidence. This
evidence is relevant to the efficiency with which the count could be conducted,
particularly in view of Mr Hatting h’s evidence that t he usual aerial counting
method is to have two counters sitting behind each other, with the pilot
determining the direction of the gr idlines, which are usually 100 to 150 metres
apart. Mr Hattingh’s evidence in this regard was consistent with that of Mr Regal,
one of Summermania’s own witnesses.
[25] Mr Hattingh raised concerns about the accuracy of aerial game counts in
general and Mr Van Niekerk’s count in particular. To substantiate this point h e
referred to an article written on the subject.2 Points of concern are set out in a
portion of the introductory summary of the article as follows:
‘For most large herbivore species, the estimates from the aerial counts were considerably lower
than those from ground counts. The data pointed to undercounting as a major problem of aerial
surveys. During the aerial counts, significant numbers of animals were missed on the transects,
first due to the low probability of spotting single animals, small groups of animals and less
conspicuous ones (sighting probability bias), and secondly because part of the population was
concealed by obstructions and therefore not visible to observers (visibility bias).’
[26] The submission on behalf of Summermania that Mr Regal responded to
these concerns is not entirely accurate. Indeed, during cross -examination Mr
Regal was asked about the ease of spot ting K udu from the air, given their
preference for dense bush. The exchange reads as follows:
‘Mr Du Toit: . . . How easy is it to spot kudu from the air?
Mr Regal: Well, depending on how much flying you’ re doing you do spot them because they
are [inaudible] they do move. You can spot them.
2 The article from which these paragraphs are drawn is titled: H Jachmann ‘Comparison of aerial counts with
ground counts for large African herbivores’ (2002) 39 published in the Journal of Applied Ecology 841-852.
12
Mr Du Toit: You can spot them, yes. I accept that. But how easy is it to spot? If you are in the
open veld and you are tracking black wildebeest, and they are running there it’s easier to spot
them than something that is standing still.
Mr Regal: They are more difficult in open plains.
Mr Du Toit: Easier to miss?
Mr Regal: Ja
Mr Du Toit: Okay. Your experience using a helicopter in game area does the helicopter scare
animals?
Mr Regal: Which is to the advantage of counting because it makes them move.
Mr Du Toit: Makes them move. Kudu, when a kudu feels threatened and it’s in the bushes does
it move or stand still?
Mr Regal: Depending on the scenario they will change, they will move or some of them will
cover. Not all of them will cover, some of them, especially the cows with young ones will
move.
Mr Du Toit: Okay. So, if there’s all of a sudden, a helicopter comes flying in and you hear the
chopping sound of the blades and the kudu is afraid of it, it might be that it will indeed, if it’s
in the dense cover that it will stay there and stand still.
Mr Regal: It is possible
Mr Du Toit: Okay. And it will be hide (sic), that you won’t not be able to spot it from the air.
Mr Regal: It is possible.’
[27] Mr Regal’s responses do not dispel the concerns expressed about aerial
counts. If anything, there is acceptance of the fact that it may be difficult to predict
how Kudus will react to instances of aerial count. The questions and answers were
general. They do not inform on how Mr Van Niekerk’s count unfolded in relation
to the aspects of concern raised about aerial counts . In addition, Mr Regal was
never qualified as a game counter in this case , although he demonstrated some
knowledge of game behaviour and game counting. The questions raised on behalf
of the Trust regarding this count remained unanswered. These relate to the game
count having been conducted by one person as opposed to the recommended two,
count having been conducted by one person as opposed to the recommended two,
the gridline spacing of 300m in this instance, rather than 100m, the terrain of the
farm not being conducive for game counting, and Mr Van Niekerk’s ability to
13
observe the speed and altitude at which the helicopter was flying whilst, at the
same time conducting the count.
[28] There was also a contention that Mr Van Niekerk had plagiarised portions
of his affidavit. He copied the following portions of his affidavit from the article
to which Mr Hattingh referred:3
‘17. The main factors that influence visibility of large herbivores from the air are animal’s
reactions to an over -flying aircraft dispersion, body size and colours. Animals that move in
response to an aircraft are more likely to be seen than static ones; dark -coloured animals are
easier to spot than light-coloured ones against a light background; large herds are easier to spot
than light-coloured ones against a light background; large heads are easier to detect than small
ones; large animals are more easily seen than small ones.
18. Body size is important while trying to spot grazers and mixed feeders from the air,
while colour is important for spotting browsers. This is mainly due to the difference in habitat
use, with browsers being confined to the thicker habitat.
19. Generally spotting and counting problems represent the most important source of bias
in a real game count. Spotting and coun ting bias may also be influenced by the density of the
vegetation, by the size and colouring of the animals, by group size, by their reaction to an over-
flying aircraft, by light conditions and by operational factors such as height and searching rate.’
[29] Indeed, these passages appear verbatim in paragraph 2 of the introductory
summary to the article. I accept that use of long-established methodologies by
experts is not unique to this case. However, the manner in which the information
is included in Mr Van Niekerk’s affidavit , without acknowledging the original
author is misleading. The content is not a mere explanation of methodology, as
was submitted on behalf o f Summermania. It is a substantive discussion of
was submitted on behalf o f Summermania. It is a substantive discussion of
general game behaviour as observed during aerial game counting. In context, and
as presented in the affidavit, it created the impression that this is information that
Mr Van Niekerk had gathered through his experience in game counting . The
3 Footnote 2 supra.
14
passages served to bolster confidence in his expertise as a knowledgeable game
counter.
[30] Even if it is accepted that from his own experience, Mr Van Niekerk agrees
with the methodology, it is ironic that the objective of the plagiarised article was
to compare the efficiency of aerial counts of large African herbivores with ground
counts. The author questioned the accuracy of aerial counts . He made the point
that estimates from aerial counts have been considerably lower than ground
counts, suggesting undercounting. He attributed this to a number of factors that
affect visibility in aerial counts, ‘even with repeat co unts of the same area’. He
posited spotting and counting problems that lead to bias (inaccuracies) in aerial
counts. These include insufficient coverage of the census area when parallel lines
are set too far apart (total count) , visual estimation of large herds, when
photography should be used; double counting as a result of poor navigation, the
quality of the observer’s eyesight and ability to concentrate for lo ng, sometimes
turbulent flights, and low probability of sighting single animals. None of these
disadvantages of aerial counting could be put to Mr Van Niekerk, and there was
no evidence in his affidavit as to how he avoided these biases during his count.
[31] Against this background , my view is that Mr Van Niekerk’s evidence
carried very little prob ative value, if any. As a result , Summermania failed to
discharge the onus to prove a breach of the warranty.
[32] Consequently the following order shall issue:
The appeal is dismissed with costs.
_______________________
N DAMBUZA
ACTING DEPUTY PRESIDENT
15
Appearances
Counsel for the appellant: S Grobler SC with him W.A Van Aswegen
Instructed by: Van Wyk Attorneys, Welkom
Hill, McHardy & Herbst Inc, Bloemfontein
Counsel for the respondents: P. S Du Toit
Instructed by: Spangenberg Attorneys, Humansdorp
Honey Attorneys, Bloemfontein