The Flechas – Insurgent Hunting in Eastern Angola 1965-1974

2013 was a great year for books in English on the Portuguese Colonial War. One of them was John P Cann’s book on the Flechas (Arrows), a specialist indigenous unit fighting for the Portuguese in Angola and later in Mozambique. The book is part of the AFRICA@WAR Series. What a find! I took a few notes.

The book is available from Amazon USA, UK, and Canada:

Cann, J. P. (2013). The Flechas. Insurgent Hunting in Eastern Angola, 1965-1974 (AFRICA@WAR Series 11)

The Flechas Cover
The Flechas Cover


April 1959

The Portuguese army established A Centre of Instruction for Special Operations (Centro de Instrucao de Operacoes Especiais or CIOE) in April 1959.

September 1959

The initial CIOE class was 480 volunteers. These were the first Special Light Infantry (Companhia de Cazadores Especiais or CCE). They were organised into a battalion of four companies. The first three companies of the battalion formed in September 1959.

April 1960

The first three CCE companies graduated from CIOE.

March 1961

UPA attacked the north.

June 1960

The fourth CCE company graduated from CIOE. Later in the month the entire battalion was deployed to Angola. They wore the first camouflage uniforms and the brown beret.

May 1961

Portuguese offensive to retake the north began in May. Progress was slow due to damage of the roads. Columns had to constantly clear felled trees, repair bridges, fill in pits and trenches. David Grenfell noted 800 felled trees and 200 trenches in the 29 km from Bungu to 31 de Janeiro. The convoys were also subject to ambush from the tall elephant grass (5-7 metres) verging the roads.

July 1961

Portuguese reoccupy Nambuangongo or ‘Nambu’ as the troops called it.

Early August

Air assaults on Quipedro.

Late August

Air assaults on Serra de Canda.

September 1961

Air assaults on Sacandia.

January 1962

The UPA/FNLA gave George Houser and John Marcum a two week tour around northern Angola. They walked 320 km over “interwoven network of trails leading through forests and elephant grass, across vine and single-log bridges, and around open-pit animal traps to the hidden semiportable village that constituted nationalist Angola” (cited p. 18). According to Marcum Portuguese air raids were “devastating” despite UPA/FNLA ploys (slit trenches as air raid shelters, mock villages as decoys, and spotters to warn of approaching aircraft).

February 1962

Batista, UPA/FNLA leader in northern Angola, was killed. He commanded an attack on the Portuguese camp at Bembe. Unbeknown to the insurgents the had been reinforced and the UPA/FNLA force was heavily outnumbered. Batista’s force was destroyed in a counterattack on 6 Feb 1962.

Early 1962

Cesare Dante Vacchi, an Italian journalist, began coaching a Portuguese battalion in the vicinity of Noqui,a frontier port on the Congo river. Vacchi had combat experience from Indochina and Algeria.

20 April 1962

Fuessi was a UPA/FNLA settlement and headquarters about 29 km east of Luvo in difficult terrain. From Fuessi mobile groups launched ambushes throughout the area.

Portuguese intelligence had identified Fuessi and the Portuguese decided to destroy it. On 20 April 1962 1st Company of Parachute Battalion 21 (120 men) was airlifted from to Luanda to Sao Salvador. Trucks transported the paratroops to Mamarrosa and then to the River Lunguege where the bridge had been destroyed. The paratroopers forded the river on foot and continued along the rough and overgrown track. It was the rainy season and even with light loads the heat and humidity made the going difficult. The paratroopers also had to advance slowly in case of ambush. Other tracks crossed the one the paratroopers were following thus confusing their guide. A Dornier observation plane flew overhead to keep them oriented and to drop supplies. The company forded the rivers Maza-Matende, Cai and Magina. The rain swollen river Luvo posed a greater challenge; the paratroopers rigged a line across the river and crossed one by one. The paratroopers reached Fuessi at noon and quietly placed counter-ambush positions surrounded the compound.

Of course the UPA / FNLA lookouts spotted the Dornier and the relatively large and slow moving column. The compound was empty. Warm fires, food, plates, utensils suggested a large group had abandoned the camp recently. Photos with French inscriptions suggested at least some of the insurgents were French speakers hence from outside Angola.

The Portuguese learnt a few lessons from this failure operation:

  • Guides had be familiar with the area of the operation
  • Smaller force sizes were better
  • Security forces had to be more comfortable in the bush
  • Spotter planes benefited the enemy as much as the security forces

Late 1962

Based on the results of Vacchi’s coaching at Noqui the Portuguese army commissioned a unit of commandos.


The first commandos were deployed in platoon sized commando groups (grupos de comandos).

September 1963

The 1st Company of Commandos began operations from Belo Horizonte. They wore the crimson beret.

Early 1965

Manuel Pontes suggests to Oscar Cardoso, a PIDE/DGS inspector, that the Bushmen could help with intelligence in eastern Angola.

Sep 1965

UPA/FNLA had two major strongholds in the Serra do Uige with the Dembos: Quivitas and CMIG-Zero. Both were tough nuts to crack being in the middle of dense mountainous jungle, on prominent hills with large, open, sweeping approaches defended by machine guns. CMIG-Zero stands for Military Centre for Instruction in Guerrilla Warfare Number Zero. CMIG-Zero included huts, a parade ground and a flag pole.

Following the Portuguese discovery of the UPA/FNLA strongholds the security forces made several failed several attempts to destroy them. It took a combined army and airforce operation in September 1965 to crack these nuts. The attack started with 20 days of artillery bombardment and air sorties. Most of the planes flew from Luanda and 16 lighter planes operated from the nearby Negage airfield (20 minutes flight time). Following this preparation the ground assault began. Several combat groups of infantry, including paratroopers, were inserted into the mountains, moved into the strongholds against light opposition, then occupied them to prevent reoccupation.

Once again the insurgents had abandoned ground rather than fight. The Portuguese had largely driven the UPA/FNLA fron the north but had not eliminated the insurgents.

Late 1965

Manuel Pontes, Oscar Cardoso, and Oscar’s wife Irene drove into the bush of Cuando Cubango to meet the Bushmen.

Early 1966

MPLA and UNITA began proselytising amongst the population of eastern Angola.

April 1966

First armed incursions in eastern Angola. The MPLA infiltrated a large group of guerrillas into the Cazombo Salient.

Oct 1966

Surge of insurgent activity near the Zambian border about 225 miles soutest of Luso.


The Portuguese began to use local auxiliaries to help intelligence.

3-12 March 1973

Operation Zeus I ran from 3-12 March 1973. The Portuguese fielded four combat groups from Infantry Battalion 4611, the 36th company of commandoes, and Flechas from Cuito Cuanavale, Mavinga and Mavengue. The aim was the clear insurgents from a triangular area about 43 miles west of Mavinga. Subaltern (Alfes) Filip Silva had his own 4th Combat Group, 3rd Company, 4611 Battalion plus a combat group of Flechas from Mavinga. Silva found the Flechas surprising. The men were issued 10 days rations which the Flechas promptly consumed in one sitting. The Flechas also made constant noise and light bright fires at night. That is until they smelt insurgents during the night. The Flechas instantly disappeared into the bush. Silva didn’t see the Flechas again but they did stay in contact by radio. Apparently they’d contacted a group of seven insurgents and were hunting them. Silva later found the bodies of two of the insurgents.

On a separate incident in 1973 Silva mentions meeting a combat group of Flechas. The Flechas had encountered insurgents the previous day. Both sides had expended their entire ammunition before withdrawing. The Flechas were now looking for a lift back to their base to refit.

Bushmen (Khum)

The whites can look but they can’t see

Bushman to Major Deville Linford in 1973, cited p. 29

The Bushmen (Khum; Bosquimanos in Portuguese) are an ethnic minority living in southern Africa. There were about 47,000 Bushmen in Africa in 1966, of which 4,700 were thought to live in Cuando Cubango in eastern Angola.

They speak dialects within the Khoisan language group so lots of clicks and pops.

An Asian ancestry is clear (at least to me). They are small (4′ 6″ on average) and very thin. They have an Asian facial structure with a broad face and tiny nose. Many have Mongolian eyefolds. They have yellow skin although this turns copper brown in the African sun. They share the curly black hair of their Negro/Bantu neighbours.

They were the original inhabitants of southern African having been there for abut 10,000 years.

The Bushmen are hunter gatherers. They have no livestock, although some groups keep dogs. They eat roots, berries, and the results of hunting. The family family groups (20-25 people) are on the move all the time looking for food.

The Bushmen were very good hunters and trackers. They would stalk an Antelope, shoot it, then track it for days until the animal died. They could recognise the individual footprints of their victim from amongst those of the wider herd, could track over hard ground, and could track at a run. They could track humans by scent, and could tell from the tracks whether the target was made by a man or woman, black or white, loaded or lightly equipped person. They could smell their Bantu opponents at some distance and could also smell landmines.

The Bushmen had a long standing enmity with the Bantu neighbours. Early on the Bushmen would sometimes immobile a Bantu insurgent with their poison arrows and then torture them to death. They would also collect enemy ears as proof of contact. The animosity was two ways and the Bantu would exterminate Bushmen camps if they found them – men, women and children. This reached genocidal proportions when Portugal lost the war in Angola.


Cardoso started with a bushman unit of eight trackers. They were armed with their traditional bows and poisoned arrows. Operations were wide ranging and made in isolation from conventional forces. Over time the level of cooperation increased although the Flechas were still at their best when tracking and fighting independently. They did however develop a special dynamic with the commandoes, each force complementing the other.

After the Bushmen trackers suffered a few losses the Portuguese began to train them as soldier.

They wore the Portuguese uniform adapted for their small size. However rather than boots they were issued canvas shoes dyed brown with coffee. Shoes and sunglasses were worn around camp but abandoned for the bush.

Flechas were issued with a variety of weapons over time as they searched for a weapon that suited their small stature:

  1. British Lee-Enfield .303 rifle. Too heavy.
  2. FN-FAL. Good in semi-automatic but Flechas couldn’t control the weapon on full-automatic.
  3. Captured AK47. Heavy but short, simple, durable, and reliable.
  4. G3. Used primarily by non-Bushmen Flechas.
  5. By 1974 the US M-16 had begun to become the standard issue.

In 1968 the Portuguese started recruiting ex-insurgents as Flechas in non-Bushman areas (Gago Coutinho, Luso and Carmona).


Oscar Cardoso

The PIDE/DGS inspector who proposed using the Bushmen as intelligence gatherers. Earlier he had held an academic chair in ethnography at the Graduate Institute of Colonial Studies. He spent seven years in Cuando Cubango.

Manuel Pontes

An administrator in Cuando Cubango who convinced Cardoso that the Bushmen would be useful to the security forces. Pontes was pivotal in recruiting the Bushmen to the Portuguese cause. Pontes lived on a raft moored on the Cuando River, on the border with Zambia. Generally the Bushmen distrusted outsiders however they respected Pontes because he had lived among them for many years and lived a modest lifestyle. The Bushmen called him Tata Khum (Father of the Khum).

Insurgent Groups

Bakongo were the main group within the UPA/FNLA. The Bakongo spanned the Angola-Congo border. The MPLA viewed the UPA/FNLA as racist (Bakongo based) and foreign (leadership based in Congo; most insurgents spoke French rather than Portuguese).

Congo-Leopoldville did not allow insurgents to carry arms within its borders. Congolese soldiers transported Congo based insurgents – whether UPA/FNLA or MPLA – to the border of Angola in trucks. Weapons were issued at the border. Upon their return to Congo the insurgents would hand in their weapons and get transported to their camps by a Congolese escort. Congo-Leopoldville used this system to discriminate against the MPLA – block their passage and confiscate their weapons.

Small groups of insurgents were inclined to surrender rather than die for the nationalist cause. Large groups for flee.


Bush terrain could vary but most was tall elephant grass. Elephant grass encroached on the roads, particularly if the roads were infrequently used.

Different types of users travelled on different paths through the forest and elephant grass. There were paths for:

  • Animals. Humans avoided these because of the insects and other dangers.
  • insurgents. And the insurgents would abandon a route once ambushed on it.
  • Refugees. Refugees avoided insurgent paths to avoid becoming impressed labour.
  • Local residents
  • Those secretly crossing the Congo-Angola border

Roads were an impediment to insurgency. They brought troops rapidly and crossing them was dangerous. Portuguese patrols we follow the tracks of those crossing the dirt roads. As a consequence both insurgents and refugees preferred stony areas to cross.

The Dembos region and the mountain ranges between Uige and Quitexe had dense forest. The Serra do Uige range is about 36 miles long by three miles wide.

The eastern zone of Angola was 420,000 square miles. More than 7.5 times the size of continental Portugal and six times the reclaimed area in the north. There were only 1.3 million people within that area, so about three per square mile. The bulk of these lived along the CFB and in the major towns. In the bush there was less than one person per square mile. The eastern area was also a plateaux about 1 km above sea level.

The Fletcha’s operational area in the eastern zone tended to be flat and featureless. The insurgents tended to camp at the headwaters of a river or at the confluence of two rivers.


Portuguese troops shipped to Angola in 1961 wore ‘canaries’. These were washed khaki and a yellowish hue.

The UPA/FNLA in 1961 wore a quasi-uniform: blue trousers with white or red bands on them.


Canhangulos: Primitive firearms made with water pipes stolen from farms.

Fletcha: Arrow in Portuguese, the primary weapon of the Bushmen hunter-gatherers.

Disclaimer: Some of the links contained within this page have my referral ID (e.g., Amazon), which provides me with a small commission for each sale. Thank you for your support

2 thoughts on “The Flechas – Insurgent Hunting in Eastern Angola 1965-1974”

  1. John P. Cann is probably one of the authors that studied more the Portuguese Colonial war and it is good to see another book from him, this time about the Flechas (Arrows).

    Even if the book is from 2013, I just knew this now, with these notes from you. So, thank you, for the news and for the interesting notes.

    Just a small note about the Portuguese. Sometimes in the text you have a typo in the word Flecha(s) and write it with an additional “t”: Fletcha(s)

    • John Cann has been quite productive over the last few years. I think I’ve got four of his books now.

      Thanks for spotting the typo.


Leave a Reply